* [dpdk-dev] [PATCH v7 0/6] Move EAL common functions. @ 2015-04-23 21:35 Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-23 21:35 UTC (permalink / raw) To: dev Changes in v7 includes Fix clang and ICC compilation errors reported by Bruce. Remove pthread _setname_ APIs introduced in v6 as a separate patch is sent. Testing: Linux - Ubuntu x86_64 14.04 Compilation successful (x86_64-native-linuxapp-gcc and x86_64-native-linuxapp-clang). "make test" results match baseline code. testpmd utility on I217/I218 Intel chipset. FreeBSD 10.0 x86_64 Compilation successful (x86_64-native-bsdapp-gcc). helloworld. Ravi Kerur (6): Move common functions in eal_thread.c Move common functions in eal.c Move common functions in eal_lcore.c Move common functions in eal_timer.c Move common functions in eal_memory.c Move common functions in eal_pci.c lib/librte_eal/bsdapp/eal/Makefile | 8 +- lib/librte_eal/bsdapp/eal/eal.c | 271 +++--------------------- lib/librte_eal/bsdapp/eal/eal_lcore.c | 72 ++----- lib/librte_eal/bsdapp/eal/eal_memory.c | 47 ++--- lib/librte_eal/bsdapp/eal/eal_pci.c | 72 +------ lib/librte_eal/bsdapp/eal/eal_thread.c | 152 -------------- lib/librte_eal/bsdapp/eal/eal_timer.c | 52 +---- lib/librte_eal/common/eal_common_app_usage.c | 63 ++++++ lib/librte_eal/common/eal_common_lcore.c | 107 ++++++++++ lib/librte_eal/common/eal_common_mem_cfg.c | 217 ++++++++++++++++++++ lib/librte_eal/common/eal_common_memory.c | 38 +++- lib/librte_eal/common/eal_common_pci.c | 72 +++++++ lib/librte_eal/common/eal_common_proc_type.c | 58 ++++++ lib/librte_eal/common/eal_common_sysfs.c | 163 +++++++++++++++ lib/librte_eal/common/eal_common_thread.c | 147 ++++++++++++- lib/librte_eal/common/eal_common_timer.c | 102 +++++++++ lib/librte_eal/common/eal_hugepages.h | 1 + lib/librte_eal/common/eal_private.h | 171 +++++++++++++++- lib/librte_eal/common/include/rte_eal.h | 4 + lib/librte_eal/linuxapp/eal/Makefile | 7 + lib/librte_eal/linuxapp/eal/eal.c | 296 ++++----------------------- lib/librte_eal/linuxapp/eal/eal_lcore.c | 66 +----- lib/librte_eal/linuxapp/eal/eal_memory.c | 36 +--- lib/librte_eal/linuxapp/eal/eal_pci.c | 75 +------ lib/librte_eal/linuxapp/eal/eal_thread.c | 152 +------------- lib/librte_eal/linuxapp/eal/eal_timer.c | 55 +---- 26 files changed, 1282 insertions(+), 1222 deletions(-) create mode 100644 lib/librte_eal/common/eal_common_app_usage.c create mode 100644 lib/librte_eal/common/eal_common_lcore.c create mode 100644 lib/librte_eal/common/eal_common_mem_cfg.c create mode 100644 lib/librte_eal/common/eal_common_proc_type.c create mode 100644 lib/librte_eal/common/eal_common_sysfs.c create mode 100644 lib/librte_eal/common/eal_common_timer.c -- 1.9.1 ^ permalink raw reply [flat|nested] 27+ messages in thread
* [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-23 21:35 [dpdk-dev] [PATCH v7 0/6] Move EAL common functions Ravi Kerur @ 2015-04-23 21:35 ` Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 2/6] Move common functions in eal.c Ravi Kerur ` (5 more replies) 0 siblings, 6 replies; 27+ messages in thread From: Ravi Kerur @ 2015-04-23 21:35 UTC (permalink / raw) To: dev Changes in v7 Remove _setname_ pthread calls. Use rte_gettid() API in RTE_LOG to print thread_id. Changes in v6 Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD respectively. Plan to use _getname_ in RTE_LOG when available. Use existing rte_get_systid() in RTE_LOG to print thread_id. Changes in v5 Rebase to latest code. Changes in v4 None Changes in v3 Changed subject to be more explicit on file name inclusion. Changes in v2 None Changes in v1 eal_thread.c has minor differences between Linux and BSD, move entire file into common directory. Use RTE_EXEC_ENV_BSDAPP to differentiate on minor differences. Rename eal_thread.c to eal_common_thread.c Makefile changes to reflect file move and name change. Fix checkpatch warnings. Signed-off-by: Ravi Kerur <rkerur@gmail.com> --- lib/librte_eal/bsdapp/eal/Makefile | 2 +- lib/librte_eal/bsdapp/eal/eal_thread.c | 152 ------------------------------ lib/librte_eal/common/eal_common_thread.c | 147 ++++++++++++++++++++++++++++- lib/librte_eal/linuxapp/eal/eal_thread.c | 152 +----------------------------- 4 files changed, 148 insertions(+), 305 deletions(-) diff --git a/lib/librte_eal/bsdapp/eal/Makefile b/lib/librte_eal/bsdapp/eal/Makefile index 2357cfa..55971b9 100644 --- a/lib/librte_eal/bsdapp/eal/Makefile +++ b/lib/librte_eal/bsdapp/eal/Makefile @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE # workaround for a gcc bug with noreturn attribute # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) -CFLAGS_eal_thread.o += -Wno-return-type +CFLAGS_eal_common_thread.o += -Wno-return-type CFLAGS_eal_hpet.o += -Wno-return-type endif diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c b/lib/librte_eal/bsdapp/eal/eal_thread.c index 9a03437..5714b8f 100644 --- a/lib/librte_eal/bsdapp/eal/eal_thread.c +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c @@ -35,163 +35,11 @@ #include <stdio.h> #include <stdlib.h> #include <stdint.h> -#include <unistd.h> -#include <sched.h> -#include <pthread_np.h> -#include <sys/queue.h> #include <sys/thr.h> -#include <rte_debug.h> -#include <rte_atomic.h> -#include <rte_launch.h> -#include <rte_log.h> -#include <rte_memory.h> -#include <rte_memzone.h> -#include <rte_per_lcore.h> -#include <rte_eal.h> -#include <rte_per_lcore.h> -#include <rte_lcore.h> - #include "eal_private.h" #include "eal_thread.h" -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; -RTE_DEFINE_PER_LCORE(unsigned, _socket_id) = (unsigned)SOCKET_ID_ANY; -RTE_DEFINE_PER_LCORE(rte_cpuset_t, _cpuset); - -/* - * Send a message to a slave lcore identified by slave_id to call a - * function f with argument arg. Once the execution is done, the - * remote lcore switch in FINISHED state. - */ -int -rte_eal_remote_launch(int (*f)(void *), void *arg, unsigned slave_id) -{ - int n; - char c = 0; - int m2s = lcore_config[slave_id].pipe_master2slave[1]; - int s2m = lcore_config[slave_id].pipe_slave2master[0]; - - if (lcore_config[slave_id].state != WAIT) - return -EBUSY; - - lcore_config[slave_id].f = f; - lcore_config[slave_id].arg = arg; - - /* send message */ - n = 0; - while (n == 0 || (n < 0 && errno == EINTR)) - n = write(m2s, &c, 1); - if (n < 0) - rte_panic("cannot write on configuration pipe\n"); - - /* wait ack */ - do { - n = read(s2m, &c, 1); - } while (n < 0 && errno == EINTR); - - if (n <= 0) - rte_panic("cannot read on configuration pipe\n"); - - return 0; -} - -/* set affinity for current thread */ -static int -eal_thread_set_affinity(void) -{ - unsigned lcore_id = rte_lcore_id(); - - /* acquire system unique id */ - rte_gettid(); - - /* update EAL thread core affinity */ - return rte_thread_set_affinity(&lcore_config[lcore_id].cpuset); -} - -void eal_thread_init_master(unsigned lcore_id) -{ - /* set the lcore ID in per-lcore memory area */ - RTE_PER_LCORE(_lcore_id) = lcore_id; - - /* set CPU affinity */ - if (eal_thread_set_affinity() < 0) - rte_panic("cannot set affinity\n"); -} - -/* main loop of threads */ -__attribute__((noreturn)) void * -eal_thread_loop(__attribute__((unused)) void *arg) -{ - char c; - int n, ret; - unsigned lcore_id; - pthread_t thread_id; - int m2s, s2m; - char cpuset[RTE_CPU_AFFINITY_STR_LEN]; - - thread_id = pthread_self(); - - /* retrieve our lcore_id from the configuration structure */ - RTE_LCORE_FOREACH_SLAVE(lcore_id) { - if (thread_id == lcore_config[lcore_id].thread_id) - break; - } - if (lcore_id == RTE_MAX_LCORE) - rte_panic("cannot retrieve lcore id\n"); - - m2s = lcore_config[lcore_id].pipe_master2slave[0]; - s2m = lcore_config[lcore_id].pipe_slave2master[1]; - - /* set the lcore ID in per-lcore memory area */ - RTE_PER_LCORE(_lcore_id) = lcore_id; - - /* set CPU affinity */ - if (eal_thread_set_affinity() < 0) - rte_panic("cannot set affinity\n"); - - ret = eal_thread_dump_affinity(cpuset, RTE_CPU_AFFINITY_STR_LEN); - - RTE_LOG(DEBUG, EAL, "lcore %u is ready (tid=%p;cpuset=[%s%s])\n", - lcore_id, thread_id, cpuset, ret == 0 ? "" : "..."); - - /* read on our pipe to get commands */ - while (1) { - void *fct_arg; - - /* wait command */ - do { - n = read(m2s, &c, 1); - } while (n < 0 && errno == EINTR); - - if (n <= 0) - rte_panic("cannot read on configuration pipe\n"); - - lcore_config[lcore_id].state = RUNNING; - - /* send ack */ - n = 0; - while (n == 0 || (n < 0 && errno == EINTR)) - n = write(s2m, &c, 1); - if (n < 0) - rte_panic("cannot write on configuration pipe\n"); - - if (lcore_config[lcore_id].f == NULL) - rte_panic("NULL function pointer\n"); - - /* call the function and store the return value */ - fct_arg = lcore_config[lcore_id].arg; - ret = lcore_config[lcore_id].f(fct_arg); - lcore_config[lcore_id].ret = ret; - rte_wmb(); - lcore_config[lcore_id].state = FINISHED; - } - - /* never reached */ - /* pthread_exit(NULL); */ - /* return NULL; */ -} - /* require calling thread tid by gettid() */ int rte_sys_gettid(void) { diff --git a/lib/librte_eal/common/eal_common_thread.c b/lib/librte_eal/common/eal_common_thread.c index 2405e93..5e55401 100644 --- a/lib/librte_eal/common/eal_common_thread.c +++ b/lib/librte_eal/common/eal_common_thread.c @@ -31,11 +31,12 @@ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ +#include <errno.h> #include <stdio.h> #include <stdlib.h> #include <stdint.h> #include <unistd.h> -#include <pthread.h> +#include <sys/queue.h> #include <sched.h> #include <assert.h> #include <string.h> @@ -43,10 +44,21 @@ #include <rte_lcore.h> #include <rte_memory.h> #include <rte_log.h> +#include <rte_debug.h> +#include <rte_atomic.h> +#include <rte_launch.h> +#include <rte_memzone.h> +#include <rte_per_lcore.h> +#include <rte_eal.h> +#include <rte_per_lcore.h> +#include "eal_private.h" #include "eal_thread.h" RTE_DECLARE_PER_LCORE(unsigned , _socket_id); +RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; +RTE_DEFINE_PER_LCORE(unsigned, _socket_id) = (unsigned)SOCKET_ID_ANY; +RTE_DEFINE_PER_LCORE(rte_cpuset_t, _cpuset); unsigned rte_socket_id(void) { @@ -155,3 +167,136 @@ exit: return ret; } + +/* + * Send a message to a slave lcore identified by slave_id to call a + * function f with argument arg. Once the execution is done, the + * remote lcore switch in FINISHED state. + */ +int +rte_eal_remote_launch(int (*f)(void *), void *arg, unsigned slave_id) +{ + int n; + char c = 0; + int m2s = lcore_config[slave_id].pipe_master2slave[1]; + int s2m = lcore_config[slave_id].pipe_slave2master[0]; + + if (lcore_config[slave_id].state != WAIT) + return -EBUSY; + + lcore_config[slave_id].f = f; + lcore_config[slave_id].arg = arg; + + /* send message */ + n = 0; + while (n == 0 || (n < 0 && errno == EINTR)) + n = write(m2s, &c, 1); + if (n < 0) + rte_panic("cannot write on configuration pipe\n"); + + /* wait ack */ + do { + n = read(s2m, &c, 1); + } while (n < 0 && errno == EINTR); + + if (n <= 0) + rte_panic("cannot read on configuration pipe\n"); + + return 0; +} + +/* set affinity for current EAL thread */ +static int +eal_thread_set_affinity(void) +{ + unsigned lcore_id = rte_lcore_id(); + + /* acquire system unique id */ + rte_gettid(); + + /* update EAL thread core affinity */ + return rte_thread_set_affinity(&lcore_config[lcore_id].cpuset); +} + +void eal_thread_init_master(unsigned lcore_id) +{ + /* set the lcore ID in per-lcore memory area */ + RTE_PER_LCORE(_lcore_id) = lcore_id; + + /* set CPU affinity */ + if (eal_thread_set_affinity() < 0) + rte_panic("cannot set affinity\n"); +} + +/* main loop of threads */ +__attribute__((noreturn)) void * +eal_thread_loop(__attribute__((unused)) void *arg) +{ + char c; + int n, ret; + unsigned lcore_id; + pthread_t thread_id; + int m2s, s2m; + char cpuset[RTE_CPU_AFFINITY_STR_LEN]; + + thread_id = pthread_self(); + + /* retrieve our lcore_id from the configuration structure */ + RTE_LCORE_FOREACH_SLAVE(lcore_id) { + if (thread_id == lcore_config[lcore_id].thread_id) + break; + } + if (lcore_id == RTE_MAX_LCORE) + rte_panic("cannot retrieve lcore id\n"); + + m2s = lcore_config[lcore_id].pipe_master2slave[0]; + s2m = lcore_config[lcore_id].pipe_slave2master[1]; + + /* set the lcore ID in per-lcore memory area */ + RTE_PER_LCORE(_lcore_id) = lcore_id; + + /* set CPU affinity */ + if (eal_thread_set_affinity() < 0) + rte_panic("cannot set affinity\n"); + + ret = eal_thread_dump_affinity(cpuset, RTE_CPU_AFFINITY_STR_LEN); + + RTE_LOG(DEBUG, EAL, "lcore %u is ready (thread=%d;cpuset=[%s%s])\n", + lcore_id, rte_gettid(), cpuset, ret == 0 ? "" : "..."); + + /* read on our pipe to get commands */ + while (1) { + void *fct_arg; + + /* wait command */ + do { + n = read(m2s, &c, 1); + } while (n < 0 && errno == EINTR); + + if (n <= 0) + rte_panic("cannot read on configuration pipe\n"); + + lcore_config[lcore_id].state = RUNNING; + + /* send ack */ + n = 0; + while (n == 0 || (n < 0 && errno == EINTR)) + n = write(s2m, &c, 1); + if (n < 0) + rte_panic("cannot write on configuration pipe\n"); + + if (lcore_config[lcore_id].f == NULL) + rte_panic("NULL function pointer\n"); + + /* call the function and store the return value */ + fct_arg = lcore_config[lcore_id].arg; + ret = lcore_config[lcore_id].f(fct_arg); + lcore_config[lcore_id].ret = ret; + rte_wmb(); + lcore_config[lcore_id].state = FINISHED; + } + + /* never reached */ + /* pthread_exit(NULL); */ + /* return NULL; */ +} diff --git a/lib/librte_eal/linuxapp/eal/eal_thread.c b/lib/librte_eal/linuxapp/eal/eal_thread.c index 18bd8e0..51dca37 100644 --- a/lib/librte_eal/linuxapp/eal/eal_thread.c +++ b/lib/librte_eal/linuxapp/eal/eal_thread.c @@ -34,163 +34,13 @@ #include <errno.h> #include <stdio.h> #include <stdlib.h> -#include <stdint.h> #include <unistd.h> -#include <pthread.h> -#include <sched.h> -#include <sys/queue.h> #include <sys/syscall.h> -#include <rte_debug.h> -#include <rte_atomic.h> -#include <rte_launch.h> -#include <rte_log.h> -#include <rte_memory.h> -#include <rte_memzone.h> -#include <rte_per_lcore.h> -#include <rte_eal.h> -#include <rte_per_lcore.h> -#include <rte_lcore.h> - #include "eal_private.h" #include "eal_thread.h" -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; -RTE_DEFINE_PER_LCORE(unsigned, _socket_id) = (unsigned)SOCKET_ID_ANY; -RTE_DEFINE_PER_LCORE(rte_cpuset_t, _cpuset); - -/* - * Send a message to a slave lcore identified by slave_id to call a - * function f with argument arg. Once the execution is done, the - * remote lcore switch in FINISHED state. - */ -int -rte_eal_remote_launch(int (*f)(void *), void *arg, unsigned slave_id) -{ - int n; - char c = 0; - int m2s = lcore_config[slave_id].pipe_master2slave[1]; - int s2m = lcore_config[slave_id].pipe_slave2master[0]; - - if (lcore_config[slave_id].state != WAIT) - return -EBUSY; - - lcore_config[slave_id].f = f; - lcore_config[slave_id].arg = arg; - - /* send message */ - n = 0; - while (n == 0 || (n < 0 && errno == EINTR)) - n = write(m2s, &c, 1); - if (n < 0) - rte_panic("cannot write on configuration pipe\n"); - - /* wait ack */ - do { - n = read(s2m, &c, 1); - } while (n < 0 && errno == EINTR); - - if (n <= 0) - rte_panic("cannot read on configuration pipe\n"); - - return 0; -} - -/* set affinity for current EAL thread */ -static int -eal_thread_set_affinity(void) -{ - unsigned lcore_id = rte_lcore_id(); - - /* acquire system unique id */ - rte_gettid(); - - /* update EAL thread core affinity */ - return rte_thread_set_affinity(&lcore_config[lcore_id].cpuset); -} - -void eal_thread_init_master(unsigned lcore_id) -{ - /* set the lcore ID in per-lcore memory area */ - RTE_PER_LCORE(_lcore_id) = lcore_id; - - /* set CPU affinity */ - if (eal_thread_set_affinity() < 0) - rte_panic("cannot set affinity\n"); -} - -/* main loop of threads */ -__attribute__((noreturn)) void * -eal_thread_loop(__attribute__((unused)) void *arg) -{ - char c; - int n, ret; - unsigned lcore_id; - pthread_t thread_id; - int m2s, s2m; - char cpuset[RTE_CPU_AFFINITY_STR_LEN]; - - thread_id = pthread_self(); - - /* retrieve our lcore_id from the configuration structure */ - RTE_LCORE_FOREACH_SLAVE(lcore_id) { - if (thread_id == lcore_config[lcore_id].thread_id) - break; - } - if (lcore_id == RTE_MAX_LCORE) - rte_panic("cannot retrieve lcore id\n"); - - m2s = lcore_config[lcore_id].pipe_master2slave[0]; - s2m = lcore_config[lcore_id].pipe_slave2master[1]; - - /* set the lcore ID in per-lcore memory area */ - RTE_PER_LCORE(_lcore_id) = lcore_id; - - /* set CPU affinity */ - if (eal_thread_set_affinity() < 0) - rte_panic("cannot set affinity\n"); - - ret = eal_thread_dump_affinity(cpuset, RTE_CPU_AFFINITY_STR_LEN); - - RTE_LOG(DEBUG, EAL, "lcore %u is ready (tid=%x;cpuset=[%s%s])\n", - lcore_id, (int)thread_id, cpuset, ret == 0 ? "" : "..."); - - /* read on our pipe to get commands */ - while (1) { - void *fct_arg; - - /* wait command */ - do { - n = read(m2s, &c, 1); - } while (n < 0 && errno == EINTR); - - if (n <= 0) - rte_panic("cannot read on configuration pipe\n"); - - lcore_config[lcore_id].state = RUNNING; - - /* send ack */ - n = 0; - while (n == 0 || (n < 0 && errno == EINTR)) - n = write(s2m, &c, 1); - if (n < 0) - rte_panic("cannot write on configuration pipe\n"); - - if (lcore_config[lcore_id].f == NULL) - rte_panic("NULL function pointer\n"); - - /* call the function and store the return value */ - fct_arg = lcore_config[lcore_id].arg; - ret = lcore_config[lcore_id].f(fct_arg); - lcore_config[lcore_id].ret = ret; - rte_wmb(); - lcore_config[lcore_id].state = FINISHED; - } - - /* never reached */ - /* pthread_exit(NULL); */ - /* return NULL; */ -} +#include <rte_log.h> /* require calling thread tid by gettid() */ int rte_sys_gettid(void) -- 1.9.1 ^ permalink raw reply [flat|nested] 27+ messages in thread
* [dpdk-dev] [PATCH v7 2/6] Move common functions in eal.c 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Ravi Kerur @ 2015-04-23 21:35 ` Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 3/6] Move common functions in eal_lcore.c Ravi Kerur ` (4 subsequent siblings) 5 siblings, 0 replies; 27+ messages in thread From: Ravi Kerur @ 2015-04-23 21:35 UTC (permalink / raw) To: dev Changes in v7 Fix clang and ICC compilation errors. Changes in v6 Split eal_common_system.c and eal_common_runtime.c into eal_common_sysfs.c eal_common_mem_cfg.c eal_common_proc_type.c eal_comm_app_usage.c based on functionality. Changes in v5 Rebase to latest code. Changes in v4 Remove eal_externs.h file, instead use _get_ and _set_ APIS to access those variables. Split eal_common.c into eal_common_system.c and and eal_common_runtime.c rte_eal prefix functions are moved to _runtime_ and eal prefix functions are moved to _system_ files respectively. Changes in v3 Changed subject to be more explicit on file name inclusion. Changes in v2 In function rte_eal_config_create remove #ifdef _BSDAPP_ and initialize mem_cfg_addr unconditionally. Changes in v1 Move common functions in eal.c to librte_eal/common/eal_common.c. Following functions are moved to eal_common.c file. struct rte_config *rte_eal_get_configuration(void); int eal_parse_sysfs_value(const char *filename, unsigned long *val); static void rte_eal_config_create(void); enum rte_proc_type_t eal_proc_type_detect(void); void rte_eal_config_init(void); rte_usage_hook_t rte_set_application_usage_hook(rte_usage_hook_t usage_func); inline size_t eal_get_hugepage_mem_size(void); void eal_check_mem_on_local_socket(void); int sync_func(__attribute__((unused)) void *arg); inline void rte_eal_mcfg_complete(void); int rte_eal_has_hugepages(void); enum rte_lcore_role_t rte_eal_lcore_role(unsigned lcore_id); enum rte_proc_type_t rte_eal_process_type(void); Makefile changes to reflect new files added. Fix checkpatch warnings and errors. Signed-off-by: Ravi Kerur <rkerur@gmail.com> --- lib/librte_eal/bsdapp/eal/Makefile | 4 + lib/librte_eal/bsdapp/eal/eal.c | 271 +++--------------------- lib/librte_eal/common/eal_common_app_usage.c | 63 ++++++ lib/librte_eal/common/eal_common_mem_cfg.c | 217 ++++++++++++++++++++ lib/librte_eal/common/eal_common_proc_type.c | 58 ++++++ lib/librte_eal/common/eal_common_sysfs.c | 163 +++++++++++++++ lib/librte_eal/common/eal_hugepages.h | 1 + lib/librte_eal/common/eal_private.h | 78 +++++++ lib/librte_eal/common/include/rte_eal.h | 4 + lib/librte_eal/linuxapp/eal/Makefile | 4 + lib/librte_eal/linuxapp/eal/eal.c | 296 ++++----------------------- 11 files changed, 668 insertions(+), 491 deletions(-) create mode 100644 lib/librte_eal/common/eal_common_app_usage.c create mode 100644 lib/librte_eal/common/eal_common_mem_cfg.c create mode 100644 lib/librte_eal/common/eal_common_proc_type.c create mode 100644 lib/librte_eal/common/eal_common_sysfs.c diff --git a/lib/librte_eal/bsdapp/eal/Makefile b/lib/librte_eal/bsdapp/eal/Makefile index 55971b9..5076a05 100644 --- a/lib/librte_eal/bsdapp/eal/Makefile +++ b/lib/librte_eal/bsdapp/eal/Makefile @@ -78,6 +78,10 @@ SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_devargs.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_dev.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_options.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_thread.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_sysfs.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_mem_cfg.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_proc_type.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_app_usage.c CFLAGS_eal.o := -D_GNU_SOURCE #CFLAGS_eal_thread.o := -D_GNU_SOURCE diff --git a/lib/librte_eal/bsdapp/eal/eal.c b/lib/librte_eal/bsdapp/eal/eal.c index 871d5f4..645500c 100644 --- a/lib/librte_eal/bsdapp/eal/eal.c +++ b/lib/librte_eal/bsdapp/eal/eal.c @@ -80,29 +80,6 @@ #include "eal_hugepages.h" #include "eal_options.h" -#define MEMSIZE_IF_NO_HUGE_PAGE (64ULL * 1024ULL * 1024ULL) - -/* Allow the application to print its usage message too if set */ -static rte_usage_hook_t rte_application_usage_hook = NULL; -/* early configuration structure, when memory config is not mmapped */ -static struct rte_mem_config early_mem_config; - -/* define fd variable here, because file needs to be kept open for the - * duration of the program, as we hold a write lock on it in the primary proc */ -static int mem_cfg_fd = -1; - -static struct flock wr_lock = { - .l_type = F_WRLCK, - .l_whence = SEEK_SET, - .l_start = offsetof(struct rte_mem_config, memseg), - .l_len = sizeof(early_mem_config.memseg), -}; - -/* Address of global and public configuration */ -static struct rte_config rte_config = { - .mem_config = &early_mem_config, -}; - /* internal configuration (per-core) */ struct lcore_config lcore_config[RTE_MAX_LCORE]; @@ -112,160 +89,57 @@ struct internal_config internal_config; /* used by rte_rdtsc() */ int rte_cycles_vmware_tsc_map; -/* Return a pointer to the configuration structure */ -struct rte_config * -rte_eal_get_configuration(void) -{ - return &rte_config; -} - -/* parse a sysfs (or other) file containing one integer value */ -int -eal_parse_sysfs_value(const char *filename, unsigned long *val) -{ - FILE *f; - char buf[BUFSIZ]; - char *end = NULL; - - if ((f = fopen(filename, "r")) == NULL) { - RTE_LOG(ERR, EAL, "%s(): cannot open sysfs value %s\n", - __func__, filename); - return -1; - } - - if (fgets(buf, sizeof(buf), f) == NULL) { - RTE_LOG(ERR, EAL, "%s(): cannot read sysfs value %s\n", - __func__, filename); - fclose(f); - return -1; - } - *val = strtoul(buf, &end, 0); - if ((buf[0] == '\0') || (end == NULL) || (*end != '\n')) { - RTE_LOG(ERR, EAL, "%s(): cannot parse sysfs value %s\n", - __func__, filename); - fclose(f); - return -1; - } - fclose(f); - return 0; -} - - -/* create memory configuration in shared/mmap memory. Take out - * a write lock on the memsegs, so we can auto-detect primary/secondary. - * This means we never close the file while running (auto-close on exit). - * We also don't lock the whole file, so that in future we can use read-locks - * on other parts, e.g. memzones, to detect if there are running secondary - * processes. */ -static void -rte_eal_config_create(void) +inline void * +rte_eal_get_mem_cfg_addr(void) { - void *rte_mem_cfg_addr; - int retval; - - const char *pathname = eal_runtime_config_path(); - - if (internal_config.no_shconf) - return; - - if (mem_cfg_fd < 0){ - mem_cfg_fd = open(pathname, O_RDWR | O_CREAT, 0660); - if (mem_cfg_fd < 0) - rte_panic("Cannot open '%s' for rte_mem_config\n", pathname); - } - - retval = ftruncate(mem_cfg_fd, sizeof(*rte_config.mem_config)); - if (retval < 0){ - close(mem_cfg_fd); - rte_panic("Cannot resize '%s' for rte_mem_config\n", pathname); - } - - retval = fcntl(mem_cfg_fd, F_SETLK, &wr_lock); - if (retval < 0){ - close(mem_cfg_fd); - rte_exit(EXIT_FAILURE, "Cannot create lock on '%s'. Is another primary " - "process running?\n", pathname); - } - - rte_mem_cfg_addr = mmap(NULL, sizeof(*rte_config.mem_config), - PROT_READ | PROT_WRITE, MAP_SHARED, mem_cfg_fd, 0); - - if (rte_mem_cfg_addr == MAP_FAILED){ - rte_panic("Cannot mmap memory for rte_config\n"); - } - memcpy(rte_mem_cfg_addr, &early_mem_config, sizeof(early_mem_config)); - rte_config.mem_config = (struct rte_mem_config *) rte_mem_cfg_addr; + return NULL; } /* attach to an existing shared memory config */ -static void +void rte_eal_config_attach(void) { - void *rte_mem_cfg_addr; + struct rte_mem_config *mem_config; + struct rte_config *rte_config; const char *pathname = eal_runtime_config_path(); + int *mem_cfg_fd = eal_get_mem_cfg_fd(); if (internal_config.no_shconf) return; - if (mem_cfg_fd < 0){ - mem_cfg_fd = open(pathname, O_RDWR); - if (mem_cfg_fd < 0) + rte_config = rte_eal_get_configuration(); + if (rte_config == NULL) + return; + + if (*mem_cfg_fd < 0) { + *mem_cfg_fd = open(pathname, O_RDWR); + if (*mem_cfg_fd < 0) rte_panic("Cannot open '%s' for rte_mem_config\n", pathname); } - rte_mem_cfg_addr = mmap(NULL, sizeof(*rte_config.mem_config), - PROT_READ | PROT_WRITE, MAP_SHARED, mem_cfg_fd, 0); - close(mem_cfg_fd); - if (rte_mem_cfg_addr == MAP_FAILED) + mem_config = (struct rte_mem_config *) mmap(NULL, sizeof(*mem_config), + PROT_READ | PROT_WRITE, + MAP_SHARED, *mem_cfg_fd, 0); + close(*mem_cfg_fd); + if (mem_config == MAP_FAILED) rte_panic("Cannot mmap memory for rte_config\n"); - rte_config.mem_config = (struct rte_mem_config *) rte_mem_cfg_addr; -} - -/* Detect if we are a primary or a secondary process */ -enum rte_proc_type_t -eal_proc_type_detect(void) -{ - enum rte_proc_type_t ptype = RTE_PROC_PRIMARY; - const char *pathname = eal_runtime_config_path(); - - /* if we can open the file but not get a write-lock we are a secondary - * process. NOTE: if we get a file handle back, we keep that open - * and don't close it to prevent a race condition between multiple opens */ - if (((mem_cfg_fd = open(pathname, O_RDWR)) >= 0) && - (fcntl(mem_cfg_fd, F_SETLK, &wr_lock) < 0)) - ptype = RTE_PROC_SECONDARY; - - RTE_LOG(INFO, EAL, "Auto-detected process type: %s\n", - ptype == RTE_PROC_PRIMARY ? "PRIMARY" : "SECONDARY"); - - return ptype; + rte_config->mem_config = mem_config; } -/* Sets up rte_config structure with the pointer to shared memory config.*/ -static void -rte_config_init(void) +/* NOP for BSD */ +void +rte_eal_config_reattach(void) { - rte_config.process_type = internal_config.process_type; - - switch (rte_config.process_type){ - case RTE_PROC_PRIMARY: - rte_eal_config_create(); - break; - case RTE_PROC_SECONDARY: - rte_eal_config_attach(); - rte_eal_mcfg_wait_complete(rte_config.mem_config); - break; - case RTE_PROC_AUTO: - case RTE_PROC_INVALID: - rte_panic("Invalid process type\n"); - } } /* display usage */ static void eal_usage(const char *prgname) { + rte_usage_hook_t rte_application_usage_hook = + rte_get_application_usage_hook(); + printf("\nUsage: %s ", prgname); eal_common_usage(); /* Allow the application to print its usage message too if hook is set */ @@ -275,37 +149,6 @@ eal_usage(const char *prgname) } } -/* Set a per-application usage message */ -rte_usage_hook_t -rte_set_application_usage_hook( rte_usage_hook_t usage_func ) -{ - rte_usage_hook_t old_func; - - /* Will be NULL on the first call to denote the last usage routine. */ - old_func = rte_application_usage_hook; - rte_application_usage_hook = usage_func; - - return old_func; -} - -static inline size_t -eal_get_hugepage_mem_size(void) -{ - uint64_t size = 0; - unsigned i, j; - - for (i = 0; i < internal_config.num_hugepage_sizes; i++) { - struct hugepage_info *hpi = &internal_config.hugepage_info[i]; - if (hpi->hugedir != NULL) { - for (j = 0; j < RTE_MAX_NUMA_NODES; j++) { - size += hpi->hugepage_sz * hpi->num_pages[j]; - } - } - } - - return (size < SIZE_MAX) ? (size_t)(size) : SIZE_MAX; -} - /* Parse the argument given in the command line of the application */ static int eal_parse_args(int argc, char **argv) @@ -378,45 +221,6 @@ eal_parse_args(int argc, char **argv) return ret; } -static void -eal_check_mem_on_local_socket(void) -{ - const struct rte_memseg *ms; - int i, socket_id; - - socket_id = rte_lcore_to_socket_id(rte_config.master_lcore); - - ms = rte_eal_get_physmem_layout(); - - for (i = 0; i < RTE_MAX_MEMSEG; i++) - if (ms[i].socket_id == socket_id && - ms[i].len > 0) - return; - - RTE_LOG(WARNING, EAL, "WARNING: Master core has no " - "memory on local socket!\n"); -} - -static int -sync_func(__attribute__((unused)) void *arg) -{ - return 0; -} - -inline static void -rte_eal_mcfg_complete(void) -{ - /* ALL shared mem_config related INIT DONE */ - if (rte_config.process_type == RTE_PROC_PRIMARY) - rte_config.mem_config->magic = RTE_MAGIC; -} - -/* return non-zero if hugepages are enabled. */ -int rte_eal_has_hugepages(void) -{ - return !internal_config.no_hugetlbfs; -} - /* Abstraction for port I/0 privilege */ int rte_eal_iopl_init(void) @@ -437,8 +241,13 @@ rte_eal_init(int argc, char **argv) int i, fctret, ret; pthread_t thread_id; static rte_atomic32_t run_once = RTE_ATOMIC32_INIT(0); + struct rte_config *rte_config; char cpuset[RTE_CPU_AFFINITY_STR_LEN]; + rte_config = rte_eal_get_configuration(); + if (rte_config == NULL) + return -1; + if (!rte_atomic32_test_and_set(&run_once)) return -1; @@ -512,12 +321,12 @@ rte_eal_init(int argc, char **argv) rte_eal_mcfg_complete(); - eal_thread_init_master(rte_config.master_lcore); + eal_thread_init_master(rte_config->master_lcore); ret = eal_thread_dump_affinity(cpuset, RTE_CPU_AFFINITY_STR_LEN); RTE_LOG(DEBUG, EAL, "Master lcore %u is ready (tid=%p;cpuset=[%s%s])\n", - rte_config.master_lcore, thread_id, cpuset, + rte_config->master_lcore, thread_id, cpuset, ret == 0 ? "" : "..."); if (rte_eal_dev_init() < 0) @@ -556,17 +365,3 @@ rte_eal_init(int argc, char **argv) return fctret; } - -/* get core role */ -enum rte_lcore_role_t -rte_eal_lcore_role(unsigned lcore_id) -{ - return (rte_config.lcore_role[lcore_id]); -} - -enum rte_proc_type_t -rte_eal_process_type(void) -{ - return (rte_config.process_type); -} - diff --git a/lib/librte_eal/common/eal_common_app_usage.c b/lib/librte_eal/common/eal_common_app_usage.c new file mode 100644 index 0000000..5f64d35 --- /dev/null +++ b/lib/librte_eal/common/eal_common_app_usage.c @@ -0,0 +1,63 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * Copyright(c) 2014 6WIND S.A. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include <stdio.h> +#include <stdlib.h> +#include <stdint.h> +#include <string.h> + +#include "eal_private.h" + +/* Allow the application to print its usage message too if set */ +rte_usage_hook_t rte_application_usage_hook = NULL; + +/* Get per-application usage message */ +rte_usage_hook_t +rte_get_application_usage_hook(void) +{ + return rte_application_usage_hook; +} + +/* Set a per-application usage message */ +rte_usage_hook_t +rte_set_application_usage_hook(rte_usage_hook_t usage_func) +{ + rte_usage_hook_t old_func; + + /* Will be NULL on the first call to denote the last usage routine. */ + old_func = rte_application_usage_hook; + rte_application_usage_hook = usage_func; + + return old_func; +} diff --git a/lib/librte_eal/common/eal_common_mem_cfg.c b/lib/librte_eal/common/eal_common_mem_cfg.c new file mode 100644 index 0000000..9bf8d92 --- /dev/null +++ b/lib/librte_eal/common/eal_common_mem_cfg.c @@ -0,0 +1,217 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * Copyright(c) 2014 6WIND S.A. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include <stdio.h> +#include <stdlib.h> +#include <stdint.h> +#include <string.h> +#include <stdarg.h> +#include <unistd.h> +#include <pthread.h> +#include <syslog.h> +#include <getopt.h> +#include <sys/file.h> +#include <stddef.h> +#include <errno.h> +#include <limits.h> +#include <errno.h> +#include <sys/mman.h> +#include <sys/queue.h> + +#include <rte_debug.h> +#include <rte_eal_memconfig.h> +#include <rte_log.h> + +#include "eal_private.h" +#include "eal_thread.h" +#include "eal_internal_cfg.h" +#include "eal_filesystem.h" +#include "eal_options.h" + +/* early configuration structure, when memory config is not mmapped */ +static struct rte_mem_config early_mem_config; + +/* define fd variable here, because file needs to be kept open for the + * duration of the program, as we hold a write lock on it in the primary proc */ +static int mem_cfg_fd = -1; + +static struct flock wr_lock = { + .l_type = F_WRLCK, + .l_whence = SEEK_SET, + .l_start = offsetof(struct rte_mem_config, memseg), + .l_len = sizeof(((struct rte_mem_config *)0)->memseg), +}; + +/* Address of global and public configuration */ +static struct rte_config rte_config = { + .mem_config = &early_mem_config, +}; + +/* Return a pointer to the configuration structure */ +struct rte_config * +rte_eal_get_configuration(void) +{ + return &rte_config; +} + +/* Return memory config file descriptor */ +int* +eal_get_mem_cfg_fd(void) +{ + return &mem_cfg_fd; +} + +/* create memory configuration in shared/mmap memory. Take out + * a write lock on the memsegs, so we can auto-detect primary/secondary. + * This means we never close the file while running (auto-close on exit). + * We also don't lock the whole file, so that in future we can use read-locks + * on other parts, e.g. memzones, to detect if there are running secondary + * processes. */ +static void +rte_eal_config_create(void) +{ + void *rte_mem_cfg_addr; + int retval; + + const char *pathname = eal_runtime_config_path(); + + if (internal_config.no_shconf) + return; + + rte_mem_cfg_addr = rte_eal_get_mem_cfg_addr(); + + if (mem_cfg_fd < 0) { + mem_cfg_fd = open(pathname, O_RDWR | O_CREAT, 0660); + if (mem_cfg_fd < 0) + rte_panic("Cannot open '%s' for rte_mem_config\n", + pathname); + } + + retval = eal_ftruncate_and_fcntl(sizeof(*rte_config.mem_config)); + + if (retval == -1) { + close(mem_cfg_fd); + rte_panic("Cannot resize '%s' for rte_mem_config\n", pathname); + } else if (retval == -2) { + close(mem_cfg_fd); + rte_exit(EXIT_FAILURE, "Cannot create lock on '%s'. " + "Is another primary process running?\n", pathname); + } + + rte_mem_cfg_addr = mmap(rte_mem_cfg_addr, + sizeof(*rte_config.mem_config), PROT_READ | PROT_WRITE, + MAP_SHARED, mem_cfg_fd, 0); + + if (rte_mem_cfg_addr == MAP_FAILED) + rte_panic("Cannot mmap memory for rte_config\n"); + + memcpy(rte_mem_cfg_addr, &early_mem_config, sizeof(early_mem_config)); + rte_config.mem_config = (struct rte_mem_config *) rte_mem_cfg_addr; + + /* store address of the config in the config itself so that secondary + * processes could later map the config into this exact location + */ + rte_config.mem_config->mem_cfg_addr = (uintptr_t) rte_mem_cfg_addr; +} + +/* Sets up rte_config structure with the pointer to shared memory config.*/ +void +rte_config_init(void) +{ + rte_config.process_type = internal_config.process_type; + + switch (rte_config.process_type) { + case RTE_PROC_PRIMARY: + rte_eal_config_create(); + break; + case RTE_PROC_SECONDARY: + rte_eal_config_attach(); + rte_eal_mcfg_wait_complete(rte_config.mem_config); + rte_eal_config_reattach(); + break; + case RTE_PROC_AUTO: + case RTE_PROC_INVALID: + rte_panic("Invalid process type\n"); + } +} + +inline void +rte_eal_mcfg_complete(void) +{ + /* ALL shared mem_config related INIT DONE */ + if (rte_config.process_type == RTE_PROC_PRIMARY) + rte_config.mem_config->magic = RTE_MAGIC; +} + +/* Detect if we are a primary or a secondary process */ +enum rte_proc_type_t +eal_proc_type_detect(void) +{ + enum rte_proc_type_t ptype = RTE_PROC_PRIMARY; + const char *pathname = eal_runtime_config_path(); + + /* if we can open the file but not get a write-lock we are + * a secondary process. NOTE: if we get a file handle back, + * we keep that open and don't close it to prevent a race + * condition between multiple opens + */ + mem_cfg_fd = open(pathname, O_RDWR); + if ((mem_cfg_fd >= 0) && + (fcntl(mem_cfg_fd, F_SETLK, &wr_lock) < 0)) + ptype = RTE_PROC_SECONDARY; + + RTE_LOG(INFO, EAL, "Auto-detected process type: %s\n", + ptype == RTE_PROC_PRIMARY ? "PRIMARY" : "SECONDARY"); + + return ptype; +} + +/* + * Perform ftruncate and fcntl operations on + * memory config file descriptor. + */ +int +eal_ftruncate_and_fcntl(size_t size) +{ + int retval; + + retval = ftruncate(mem_cfg_fd, size); + if (retval < 0) + return -1; + + retval = fcntl(mem_cfg_fd, F_SETLK, &wr_lock); + if (retval < 0) + return -2; + return 0; +} diff --git a/lib/librte_eal/common/eal_common_proc_type.c b/lib/librte_eal/common/eal_common_proc_type.c new file mode 100644 index 0000000..f8bb47f --- /dev/null +++ b/lib/librte_eal/common/eal_common_proc_type.c @@ -0,0 +1,58 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * Copyright(c) 2014 6WIND S.A. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include <stdio.h> +#include <stdlib.h> +#include <stdint.h> +#include <string.h> +#include <stdarg.h> +#include <unistd.h> + +#include "eal_private.h" + +#include <rte_log.h> + +enum rte_proc_type_t +rte_eal_process_type(void) +{ + struct rte_config *rte_config = + rte_eal_get_configuration(); + + if (rte_config == NULL) { + RTE_LOG(WARNING, EAL, "WARNING: rte_config NULL!\n"); + return RTE_PROC_INVALID; + } + + return rte_config->process_type; +} diff --git a/lib/librte_eal/common/eal_common_sysfs.c b/lib/librte_eal/common/eal_common_sysfs.c new file mode 100644 index 0000000..34ea3d2 --- /dev/null +++ b/lib/librte_eal/common/eal_common_sysfs.c @@ -0,0 +1,163 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * Copyright(c) 2014 6WIND S.A. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include <stdio.h> +#include <stdlib.h> +#include <stdint.h> +#include <string.h> +#include <stdarg.h> +#include <unistd.h> +#include <pthread.h> +#include <syslog.h> +#include <getopt.h> +#include <sys/file.h> +#include <stddef.h> +#include <errno.h> +#include <limits.h> +#include <errno.h> +#include <sys/mman.h> +#include <sys/queue.h> + +#include "eal_private.h" +#include "eal_internal_cfg.h" +#include "eal_filesystem.h" +#include "eal_hugepages.h" +#include "eal_options.h" + +#include <rte_log.h> +#include <rte_memory.h> +#include <rte_lcore.h> + +/* parse a sysfs (or other) file containing one integer value */ +int +eal_parse_sysfs_value(const char *filename, unsigned long *val) +{ + FILE *f; + char buf[BUFSIZ]; + char *end = NULL; + + f = fopen(filename, "r"); + if (f == NULL) { + RTE_LOG(ERR, EAL, "%s(): cannot open sysfs value %s\n", + __func__, filename); + return -1; + } + + if (fgets(buf, sizeof(buf), f) == NULL) { + RTE_LOG(ERR, EAL, "%s(): cannot read sysfs value %s\n", + __func__, filename); + fclose(f); + return -1; + } + *val = strtoul(buf, &end, 0); + if ((buf[0] == '\0') || (end == NULL) || (*end != '\n')) { + RTE_LOG(ERR, EAL, "%s(): cannot parse sysfs value %s\n", + __func__, filename); + fclose(f); + return -1; + } + fclose(f); + return 0; +} + +inline size_t +eal_get_hugepage_mem_size(void) +{ + uint64_t size = 0; + unsigned i, j; + + for (i = 0; i < internal_config.num_hugepage_sizes; i++) { + struct hugepage_info *hpi = &internal_config.hugepage_info[i]; + + if (hpi->hugedir != NULL) { + for (j = 0; j < RTE_MAX_NUMA_NODES; j++) + size += hpi->hugepage_sz * hpi->num_pages[j]; + } + } + + return (size < SIZE_MAX) ? (size_t)(size) : SIZE_MAX; +} + +void +eal_check_mem_on_local_socket(void) +{ + const struct rte_memseg *ms; + int i, socket_id; + struct rte_config *rte_config = + rte_eal_get_configuration(); + + if (rte_config == NULL) { + RTE_LOG(WARNING, EAL, "WARNING: rte_config NULL!\n"); + return; + } + + socket_id = rte_lcore_to_socket_id(rte_config->master_lcore); + + ms = rte_eal_get_physmem_layout(); + + for (i = 0; i < RTE_MAX_MEMSEG; i++) + if (ms[i].socket_id == socket_id && + ms[i].len > 0) + return; + + RTE_LOG(WARNING, EAL, "WARNING: Master core has no " + "memory on local socket!\n"); +} + +int +sync_func(__attribute__((unused)) void *arg) +{ + return 0; +} + +/* return non-zero if hugepages are enabled. */ +int rte_eal_has_hugepages(void) +{ + return !internal_config.no_hugetlbfs; +} + +/* get core role */ +enum rte_lcore_role_t +rte_eal_lcore_role(unsigned lcore_id) +{ + struct rte_config *rte_config = + rte_eal_get_configuration(); + + if (rte_config == NULL) { + RTE_LOG(WARNING, EAL, "WARNING: rte_config NULL!\n"); + return ROLE_OFF; + } + + return rte_config->lcore_role[lcore_id]; +} diff --git a/lib/librte_eal/common/eal_hugepages.h b/lib/librte_eal/common/eal_hugepages.h index 38edac0..d79ef8a 100644 --- a/lib/librte_eal/common/eal_hugepages.h +++ b/lib/librte_eal/common/eal_hugepages.h @@ -63,5 +63,6 @@ struct hugepage_file { * for the EAL to use */ int eal_hugepage_info_init(void); +size_t eal_get_hugepage_mem_size(void); #endif /* EAL_HUGEPAGES_H */ diff --git a/lib/librte_eal/common/eal_private.h b/lib/librte_eal/common/eal_private.h index 4acf5a0..bcf603f 100644 --- a/lib/librte_eal/common/eal_private.h +++ b/lib/librte_eal/common/eal_private.h @@ -36,6 +36,8 @@ #include <stdio.h> +#include <rte_eal.h> + /** * Initialize the memzone subsystem (private to eal). * @@ -232,4 +234,80 @@ int rte_eal_dev_init(void); */ int rte_eal_check_module(const char *module_name); +/** + * This function sets up rte_config structure + * + * This function is private to the EAL. + */ +void rte_config_init(void); + +/** + * This function checks memory on local socket(NUMA) + * + * This function is private to the EAL. + */ +void eal_check_mem_on_local_socket(void); + +/** + * This function updates shared mem_config INIT DONE + * + * This function is private to the EAL. + */ +void rte_eal_mcfg_complete(void); + +/** + * + * This function is private to the EAL. + */ +int sync_func(__attribute__((unused)) void *arg); + +/** + * + * This function is private to the EAL. + */ +void *rte_eal_get_mem_cfg_addr(void); + +/** + * Return a pointer to the configuration structure + * + * This function is private to the EAL. + */ +struct rte_config *rte_eal_get_configuration(void); + +/** + * Return memory config file descriptor + * + * This function is private to the EAL. + */ +int *eal_get_mem_cfg_fd(void); + +/** + * Perform ftruncate and fcntl operations on + * memory config file descriptor. + * + * This function is private to the EAL. + */ +int eal_ftruncate_and_fcntl(size_t size); + +/** + * Get per-application usage message + * + * This function is private to the EAL. + */ +rte_usage_hook_t rte_get_application_usage_hook(void); + +/** + * This function attaches shared memory config + * + * This function is private to the EAL. + */ +void rte_eal_config_attach(void); + +/** + * This function reattaches shared memory config + * + * This function is private to the EAL. + */ +void rte_eal_config_reattach(void); + #endif /* _EAL_PRIVATE_H_ */ diff --git a/lib/librte_eal/common/include/rte_eal.h b/lib/librte_eal/common/include/rte_eal.h index 1385a73..daf2ee0 100644 --- a/lib/librte_eal/common/include/rte_eal.h +++ b/lib/librte_eal/common/include/rte_eal.h @@ -51,6 +51,10 @@ extern "C" { #define RTE_MAGIC 19820526 /**< Magic number written by the main partition when ready. */ +#define MEMSIZE_IF_NO_HUGE_PAGE (64ULL * 1024ULL * 1024ULL) + +#define SOCKET_MEM_STRLEN (RTE_MAX_NUMA_NODES * 10) + /** * The lcore role (used in RTE or not). */ diff --git a/lib/librte_eal/linuxapp/eal/Makefile b/lib/librte_eal/linuxapp/eal/Makefile index 01f7b70..329d9e4 100644 --- a/lib/librte_eal/linuxapp/eal/Makefile +++ b/lib/librte_eal/linuxapp/eal/Makefile @@ -90,6 +90,10 @@ SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_devargs.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_dev.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_options.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_thread.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c CFLAGS_eal.o := -D_GNU_SOURCE CFLAGS_eal_interrupts.o := -D_GNU_SOURCE diff --git a/lib/librte_eal/linuxapp/eal/eal.c b/lib/librte_eal/linuxapp/eal/eal.c index bd770cf..a7de8e0 100644 --- a/lib/librte_eal/linuxapp/eal/eal.c +++ b/lib/librte_eal/linuxapp/eal/eal.c @@ -84,13 +84,6 @@ #include "eal_hugepages.h" #include "eal_options.h" -#define MEMSIZE_IF_NO_HUGE_PAGE (64ULL * 1024ULL * 1024ULL) - -#define SOCKET_MEM_STRLEN (RTE_MAX_NUMA_NODES * 10) - -/* Allow the application to print its usage message too if set */ -static rte_usage_hook_t rte_application_usage_hook = NULL; - TAILQ_HEAD(shared_driver_list, shared_driver); /* Definition for shared object drivers. */ @@ -105,25 +98,6 @@ struct shared_driver { static struct shared_driver_list solib_list = TAILQ_HEAD_INITIALIZER(solib_list); -/* early configuration structure, when memory config is not mmapped */ -static struct rte_mem_config early_mem_config; - -/* define fd variable here, because file needs to be kept open for the - * duration of the program, as we hold a write lock on it in the primary proc */ -static int mem_cfg_fd = -1; - -static struct flock wr_lock = { - .l_type = F_WRLCK, - .l_whence = SEEK_SET, - .l_start = offsetof(struct rte_mem_config, memseg), - .l_len = sizeof(early_mem_config.memseg), -}; - -/* Address of global and public configuration */ -static struct rte_config rte_config = { - .mem_config = &early_mem_config, -}; - /* internal configuration (per-core) */ struct lcore_config lcore_config[RTE_MAX_LCORE]; @@ -133,196 +107,85 @@ struct internal_config internal_config; /* used by rte_rdtsc() */ int rte_cycles_vmware_tsc_map; -/* Return a pointer to the configuration structure */ -struct rte_config * -rte_eal_get_configuration(void) -{ - return &rte_config; -} - -/* parse a sysfs (or other) file containing one integer value */ -int -eal_parse_sysfs_value(const char *filename, unsigned long *val) -{ - FILE *f; - char buf[BUFSIZ]; - char *end = NULL; - - if ((f = fopen(filename, "r")) == NULL) { - RTE_LOG(ERR, EAL, "%s(): cannot open sysfs value %s\n", - __func__, filename); - return -1; - } - - if (fgets(buf, sizeof(buf), f) == NULL) { - RTE_LOG(ERR, EAL, "%s(): cannot read sysfs value %s\n", - __func__, filename); - fclose(f); - return -1; - } - *val = strtoul(buf, &end, 0); - if ((buf[0] == '\0') || (end == NULL) || (*end != '\n')) { - RTE_LOG(ERR, EAL, "%s(): cannot parse sysfs value %s\n", - __func__, filename); - fclose(f); - return -1; - } - fclose(f); - return 0; -} - - -/* create memory configuration in shared/mmap memory. Take out - * a write lock on the memsegs, so we can auto-detect primary/secondary. - * This means we never close the file while running (auto-close on exit). - * We also don't lock the whole file, so that in future we can use read-locks - * on other parts, e.g. memzones, to detect if there are running secondary - * processes. */ -static void -rte_eal_config_create(void) +inline void * +rte_eal_get_mem_cfg_addr(void) { - void *rte_mem_cfg_addr; - int retval; - - const char *pathname = eal_runtime_config_path(); - - if (internal_config.no_shconf) - return; + void *mem_cfg_addr; - /* map the config before hugepage address so that we don't waste a page */ if (internal_config.base_virtaddr != 0) - rte_mem_cfg_addr = (void *) + mem_cfg_addr = (void *) RTE_ALIGN_FLOOR(internal_config.base_virtaddr - sizeof(struct rte_mem_config), sysconf(_SC_PAGE_SIZE)); else - rte_mem_cfg_addr = NULL; - - if (mem_cfg_fd < 0){ - mem_cfg_fd = open(pathname, O_RDWR | O_CREAT, 0660); - if (mem_cfg_fd < 0) - rte_panic("Cannot open '%s' for rte_mem_config\n", pathname); - } - - retval = ftruncate(mem_cfg_fd, sizeof(*rte_config.mem_config)); - if (retval < 0){ - close(mem_cfg_fd); - rte_panic("Cannot resize '%s' for rte_mem_config\n", pathname); - } - - retval = fcntl(mem_cfg_fd, F_SETLK, &wr_lock); - if (retval < 0){ - close(mem_cfg_fd); - rte_exit(EXIT_FAILURE, "Cannot create lock on '%s'. Is another primary " - "process running?\n", pathname); - } - - rte_mem_cfg_addr = mmap(rte_mem_cfg_addr, sizeof(*rte_config.mem_config), - PROT_READ | PROT_WRITE, MAP_SHARED, mem_cfg_fd, 0); - - if (rte_mem_cfg_addr == MAP_FAILED){ - rte_panic("Cannot mmap memory for rte_config\n"); - } - memcpy(rte_mem_cfg_addr, &early_mem_config, sizeof(early_mem_config)); - rte_config.mem_config = (struct rte_mem_config *) rte_mem_cfg_addr; - - /* store address of the config in the config itself so that secondary - * processes could later map the config into this exact location */ - rte_config.mem_config->mem_cfg_addr = (uintptr_t) rte_mem_cfg_addr; + mem_cfg_addr = NULL; + return mem_cfg_addr; } /* attach to an existing shared memory config */ -static void +void rte_eal_config_attach(void) { struct rte_mem_config *mem_config; + struct rte_config *rte_config; + int *mem_cfg_fd = eal_get_mem_cfg_fd(); const char *pathname = eal_runtime_config_path(); if (internal_config.no_shconf) return; - if (mem_cfg_fd < 0){ - mem_cfg_fd = open(pathname, O_RDWR); - if (mem_cfg_fd < 0) + rte_config = rte_eal_get_configuration(); + if (rte_config == NULL) + return; + + if (*mem_cfg_fd < 0) { + *mem_cfg_fd = open(pathname, O_RDWR); + if (*mem_cfg_fd < 0) rte_panic("Cannot open '%s' for rte_mem_config\n", pathname); } /* map it as read-only first */ mem_config = (struct rte_mem_config *) mmap(NULL, sizeof(*mem_config), - PROT_READ, MAP_SHARED, mem_cfg_fd, 0); + PROT_READ, MAP_SHARED, *mem_cfg_fd, 0); if (mem_config == MAP_FAILED) rte_panic("Cannot mmap memory for rte_config\n"); - rte_config.mem_config = mem_config; + rte_config->mem_config = mem_config; } /* reattach the shared config at exact memory location primary process has it */ -static void +void rte_eal_config_reattach(void) { struct rte_mem_config *mem_config; void *rte_mem_cfg_addr; + struct rte_config *rte_config; + int *mem_cfg_fd = eal_get_mem_cfg_fd(); if (internal_config.no_shconf) return; + rte_config = rte_eal_get_configuration(); + if (rte_config == NULL) + return; + /* save the address primary process has mapped shared config to */ - rte_mem_cfg_addr = (void *) (uintptr_t) rte_config.mem_config->mem_cfg_addr; + rte_mem_cfg_addr = + (void *) (uintptr_t) rte_config->mem_config->mem_cfg_addr; /* unmap original config */ - munmap(rte_config.mem_config, sizeof(struct rte_mem_config)); + munmap(rte_config->mem_config, sizeof(struct rte_mem_config)); /* remap the config at proper address */ mem_config = (struct rte_mem_config *) mmap(rte_mem_cfg_addr, - sizeof(*mem_config), PROT_READ | PROT_WRITE, MAP_SHARED, - mem_cfg_fd, 0); - close(mem_cfg_fd); + sizeof(*mem_config), PROT_READ | PROT_WRITE, + MAP_SHARED, *mem_cfg_fd, 0); + close(*mem_cfg_fd); if (mem_config == MAP_FAILED || mem_config != rte_mem_cfg_addr) rte_panic("Cannot mmap memory for rte_config\n"); - rte_config.mem_config = mem_config; -} - -/* Detect if we are a primary or a secondary process */ -enum rte_proc_type_t -eal_proc_type_detect(void) -{ - enum rte_proc_type_t ptype = RTE_PROC_PRIMARY; - const char *pathname = eal_runtime_config_path(); - - /* if we can open the file but not get a write-lock we are a secondary - * process. NOTE: if we get a file handle back, we keep that open - * and don't close it to prevent a race condition between multiple opens */ - if (((mem_cfg_fd = open(pathname, O_RDWR)) >= 0) && - (fcntl(mem_cfg_fd, F_SETLK, &wr_lock) < 0)) - ptype = RTE_PROC_SECONDARY; - - RTE_LOG(INFO, EAL, "Auto-detected process type: %s\n", - ptype == RTE_PROC_PRIMARY ? "PRIMARY" : "SECONDARY"); - - return ptype; -} - -/* Sets up rte_config structure with the pointer to shared memory config.*/ -static void -rte_config_init(void) -{ - rte_config.process_type = internal_config.process_type; - - switch (rte_config.process_type){ - case RTE_PROC_PRIMARY: - rte_eal_config_create(); - break; - case RTE_PROC_SECONDARY: - rte_eal_config_attach(); - rte_eal_mcfg_wait_complete(rte_config.mem_config); - rte_eal_config_reattach(); - break; - case RTE_PROC_AUTO: - case RTE_PROC_INVALID: - rte_panic("Invalid process type\n"); - } + rte_config->mem_config = mem_config; } /* Unlocks hugepage directories that were locked by eal_hugepage_info_init */ @@ -348,6 +211,9 @@ eal_hugedirs_unlock(void) static void eal_usage(const char *prgname) { + rte_usage_hook_t rte_application_usage_hook = + rte_get_application_usage_hook(); + printf("\nUsage: %s ", prgname); eal_common_usage(); printf("EAL Linux options:\n" @@ -367,19 +233,6 @@ eal_usage(const char *prgname) } } -/* Set a per-application usage message */ -rte_usage_hook_t -rte_set_application_usage_hook( rte_usage_hook_t usage_func ) -{ - rte_usage_hook_t old_func; - - /* Will be NULL on the first call to denote the last usage routine. */ - old_func = rte_application_usage_hook; - rte_application_usage_hook = usage_func; - - return old_func; -} - static int eal_parse_socket_mem(char *socket_mem) { @@ -481,24 +334,6 @@ eal_parse_vfio_intr(const char *mode) return -1; } -static inline size_t -eal_get_hugepage_mem_size(void) -{ - uint64_t size = 0; - unsigned i, j; - - for (i = 0; i < internal_config.num_hugepage_sizes; i++) { - struct hugepage_info *hpi = &internal_config.hugepage_info[i]; - if (hpi->hugedir != NULL) { - for (j = 0; j < RTE_MAX_NUMA_NODES; j++) { - size += hpi->hugepage_sz * hpi->num_pages[j]; - } - } - } - - return (size < SIZE_MAX) ? (size_t)(size) : SIZE_MAX; -} - /* Parse the argument given in the command line of the application */ static int eal_parse_args(int argc, char **argv) @@ -645,39 +480,6 @@ eal_parse_args(int argc, char **argv) return ret; } -static void -eal_check_mem_on_local_socket(void) -{ - const struct rte_memseg *ms; - int i, socket_id; - - socket_id = rte_lcore_to_socket_id(rte_config.master_lcore); - - ms = rte_eal_get_physmem_layout(); - - for (i = 0; i < RTE_MAX_MEMSEG; i++) - if (ms[i].socket_id == socket_id && - ms[i].len > 0) - return; - - RTE_LOG(WARNING, EAL, "WARNING: Master core has no " - "memory on local socket!\n"); -} - -static int -sync_func(__attribute__((unused)) void *arg) -{ - return 0; -} - -inline static void -rte_eal_mcfg_complete(void) -{ - /* ALL shared mem_config related INIT DONE */ - if (rte_config.process_type == RTE_PROC_PRIMARY) - rte_config.mem_config->magic = RTE_MAGIC; -} - /* * Request iopl privilege for all RPL, returns 0 on success * iopl() call is mostly for the i386 architecture. For other architectures, @@ -706,6 +508,12 @@ rte_eal_init(int argc, char **argv) const char *logid; char cpuset[RTE_CPU_AFFINITY_STR_LEN]; + struct rte_config *rte_config; + + rte_config = rte_eal_get_configuration(); + if (rte_config == NULL) + return -1; + if (!rte_atomic32_test_and_set(&run_once)) return -1; @@ -803,12 +611,12 @@ rte_eal_init(int argc, char **argv) RTE_LOG(WARNING, EAL, "%s\n", dlerror()); } - eal_thread_init_master(rte_config.master_lcore); + eal_thread_init_master(rte_config->master_lcore); ret = eal_thread_dump_affinity(cpuset, RTE_CPU_AFFINITY_STR_LEN); RTE_LOG(DEBUG, EAL, "Master lcore %u is ready (tid=%x;cpuset=[%s%s])\n", - rte_config.master_lcore, (int)thread_id, cpuset, + rte_config->master_lcore, (int)thread_id, cpuset, ret == 0 ? "" : "..."); if (rte_eal_dev_init() < 0) @@ -848,24 +656,6 @@ rte_eal_init(int argc, char **argv) return fctret; } -/* get core role */ -enum rte_lcore_role_t -rte_eal_lcore_role(unsigned lcore_id) -{ - return (rte_config.lcore_role[lcore_id]); -} - -enum rte_proc_type_t -rte_eal_process_type(void) -{ - return (rte_config.process_type); -} - -int rte_eal_has_hugepages(void) -{ - return ! internal_config.no_hugetlbfs; -} - int rte_eal_check_module(const char *module_name) { -- 1.9.1 ^ permalink raw reply [flat|nested] 27+ messages in thread
* [dpdk-dev] [PATCH v7 3/6] Move common functions in eal_lcore.c 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 2/6] Move common functions in eal.c Ravi Kerur @ 2015-04-23 21:35 ` Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 4/6] Move common functions in eal_timer.c Ravi Kerur ` (3 subsequent siblings) 5 siblings, 0 replies; 27+ messages in thread From: Ravi Kerur @ 2015-04-23 21:35 UTC (permalink / raw) To: dev Changes in v7 None Changes in v6 None Changes in v5 Rebase to latest code. Changes in v4 Implement cpu_detected() for BSD. Have common RTE_LOG for Linux and BSD in rte_eal_cpu_init(). Remove RTE_EXEC_ENV_BSDAPP in common file. Changes in v3 Changed subject to be more explicit on file name inclusion. Changes in v2 None Changes in v1 Move common function in eal_lcore.c to librte_eal/common/ eal_common_lcore.c file. Following function is moved to eal_common_lcore.c file int rte_eal_cpu_init(void); Use RTE_EXEC_ENV_BSDAPP to differentiate minor differences in common function. Makefile changes to reflect new file added. Fix checkpatch warnings and errors. Signed-off-by: Ravi Kerur <rkerur@gmail.com> --- lib/librte_eal/bsdapp/eal/Makefile | 1 + lib/librte_eal/bsdapp/eal/eal_lcore.c | 72 +++++---------------- lib/librte_eal/common/eal_common_lcore.c | 107 +++++++++++++++++++++++++++++++ lib/librte_eal/common/eal_private.h | 14 ++++ lib/librte_eal/linuxapp/eal/Makefile | 2 + lib/librte_eal/linuxapp/eal/eal_lcore.c | 66 ++----------------- 6 files changed, 143 insertions(+), 119 deletions(-) create mode 100644 lib/librte_eal/common/eal_common_lcore.c diff --git a/lib/librte_eal/bsdapp/eal/Makefile b/lib/librte_eal/bsdapp/eal/Makefile index 5076a05..1da0410 100644 --- a/lib/librte_eal/bsdapp/eal/Makefile +++ b/lib/librte_eal/bsdapp/eal/Makefile @@ -82,6 +82,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_sysfs.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_mem_cfg.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_proc_type.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_app_usage.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_lcore.c CFLAGS_eal.o := -D_GNU_SOURCE #CFLAGS_eal_thread.o := -D_GNU_SOURCE diff --git a/lib/librte_eal/bsdapp/eal/eal_lcore.c b/lib/librte_eal/bsdapp/eal/eal_lcore.c index 162fb4f..b47eb1b 100644 --- a/lib/librte_eal/bsdapp/eal/eal_lcore.c +++ b/lib/librte_eal/bsdapp/eal/eal_lcore.c @@ -44,11 +44,14 @@ #include "eal_thread.h" /* No topology information available on FreeBSD including NUMA info */ -#define cpu_core_id(X) 0 -#define cpu_socket_id(X) 0 +unsigned +eal_cpu_core_id(__rte_unused unsigned lcore_id) +{ + return 0; +} static int -get_ncpus(void) +eal_get_ncpus(void) { int mib[2] = {CTL_HW, HW_NCPU}; int ncpu; @@ -59,63 +62,18 @@ get_ncpus(void) return ncpu; } -/* - * fill the cpu_info structure with as much info as we can get. - * code is similar to linux version, but sadly available info is less. - */ -int -rte_eal_cpu_init(void) +unsigned +eal_cpu_socket_id(__rte_unused unsigned cpu_id) { - /* pointer to global configuration */ - struct rte_config *config = rte_eal_get_configuration(); - unsigned lcore_id; - unsigned count = 0; - - const unsigned ncpus = get_ncpus(); - /* - * Parse the maximum set of logical cores, detect the subset of running - * ones and enable them by default. - */ - for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { - /* init cpuset for per lcore config */ - CPU_ZERO(&lcore_config[lcore_id].cpuset); - - lcore_config[lcore_id].detected = (lcore_id < ncpus); - if (lcore_config[lcore_id].detected == 0) { - config->lcore_role[lcore_id] = ROLE_OFF; - continue; - } - - /* By default, lcore 1:1 map to cpu id */ - CPU_SET(lcore_id, &lcore_config[lcore_id].cpuset); - - /* By default, each detected core is enabled */ - config->lcore_role[lcore_id] = ROLE_RTE; - lcore_config[lcore_id].core_id = cpu_core_id(lcore_id); - lcore_config[lcore_id].socket_id = cpu_socket_id(lcore_id); - if (lcore_config[lcore_id].socket_id >= RTE_MAX_NUMA_NODES) -#ifdef RTE_EAL_ALLOW_INV_SOCKET_ID - lcore_config[lcore_id].socket_id = 0; -#else - rte_panic("Socket ID (%u) is greater than " - "RTE_MAX_NUMA_NODES (%d)\n", - lcore_config[lcore_id].socket_id, RTE_MAX_NUMA_NODES); -#endif - RTE_LOG(DEBUG, EAL, "Detected lcore %u\n", - lcore_id); - count++; - } - /* Set the count of enabled logical cores of the EAL configuration */ - config->lcore_count = count; - RTE_LOG(DEBUG, EAL, "Support maximum %u logical core(s) by configuration.\n", - RTE_MAX_LCORE); - RTE_LOG(DEBUG, EAL, "Detected %u lcore(s)\n", config->lcore_count); - return 0; } -unsigned -eal_cpu_socket_id(__rte_unused unsigned cpu_id) +/* Check if a cpu is present by the presence of the + * cpu information for it. + */ +int +eal_cpu_detected(unsigned lcore_id) { - return cpu_socket_id(cpu_id); + const unsigned ncpus = eal_get_ncpus(); + return (lcore_id < ncpus); } diff --git a/lib/librte_eal/common/eal_common_lcore.c b/lib/librte_eal/common/eal_common_lcore.c new file mode 100644 index 0000000..845140b --- /dev/null +++ b/lib/librte_eal/common/eal_common_lcore.c @@ -0,0 +1,107 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include <unistd.h> +#include <limits.h> +#include <string.h> +#include <dirent.h> + +#include <rte_log.h> +#include <rte_eal.h> +#include <rte_lcore.h> +#include <rte_common.h> +#include <rte_debug.h> + +#include "eal_private.h" +#include "eal_thread.h" + +/* + * Parse /sys/devices/system/cpu to get the number of physical and logical + * processors on the machine. The function will fill the cpu_info + * structure. + */ +int +rte_eal_cpu_init(void) +{ + /* pointer to global configuration */ + struct rte_config *config = rte_eal_get_configuration(); + unsigned lcore_id; + unsigned count = 0; + + /* + * Parse the maximum set of logical cores, detect the subset of running + * ones and enable them by default. + */ + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { + /* init cpuset for per lcore config */ + CPU_ZERO(&lcore_config[lcore_id].cpuset); + + /* in 1:1 mapping, record related cpu detected state */ + lcore_config[lcore_id].detected = eal_cpu_detected(lcore_id); + if (lcore_config[lcore_id].detected == 0) { + config->lcore_role[lcore_id] = ROLE_OFF; + continue; + } + + /* By default, lcore 1:1 map to cpu id */ + CPU_SET(lcore_id, &lcore_config[lcore_id].cpuset); + + /* By default, each detected core is enabled */ + config->lcore_role[lcore_id] = ROLE_RTE; + lcore_config[lcore_id].core_id = eal_cpu_core_id(lcore_id); + lcore_config[lcore_id].socket_id = eal_cpu_socket_id(lcore_id); + if (lcore_config[lcore_id].socket_id >= RTE_MAX_NUMA_NODES) +#ifdef RTE_EAL_ALLOW_INV_SOCKET_ID + lcore_config[lcore_id].socket_id = 0; +#else + rte_panic("Socket ID (%u) is greater than " + "RTE_MAX_NUMA_NODES (%d)\n", + lcore_config[lcore_id].socket_id, + RTE_MAX_NUMA_NODES); +#endif + + RTE_LOG(DEBUG, EAL, "Detected lcore %u as " + "core %u on socket %u\n", + lcore_id, lcore_config[lcore_id].core_id, + lcore_config[lcore_id].socket_id); + count++; + } + /* Set the count of enabled logical cores of the EAL configuration */ + config->lcore_count = count; + RTE_LOG(DEBUG, EAL, + "Support maximum %u logical core(s) by configuration.\n", + RTE_MAX_LCORE); + RTE_LOG(DEBUG, EAL, "Detected %u lcore(s)\n", config->lcore_count); + + return 0; +} diff --git a/lib/librte_eal/common/eal_private.h b/lib/librte_eal/common/eal_private.h index bcf603f..76e47a9 100644 --- a/lib/librte_eal/common/eal_private.h +++ b/lib/librte_eal/common/eal_private.h @@ -310,4 +310,18 @@ void rte_eal_config_attach(void); */ void rte_eal_config_reattach(void); +/** + * This function gets cpu core_id + * + * This function is private to the EAL. + */ +unsigned eal_cpu_core_id(unsigned lcore_id); + +/** + * This function check if cpu is present + * + * This function is private to the EAL. + */ +int eal_cpu_detected(unsigned lcore_id); + #endif /* _EAL_PRIVATE_H_ */ diff --git a/lib/librte_eal/linuxapp/eal/Makefile b/lib/librte_eal/linuxapp/eal/Makefile index 329d9e4..7259bd7 100644 --- a/lib/librte_eal/linuxapp/eal/Makefile +++ b/lib/librte_eal/linuxapp/eal/Makefile @@ -94,6 +94,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_lcore.c CFLAGS_eal.o := -D_GNU_SOURCE CFLAGS_eal_interrupts.o := -D_GNU_SOURCE @@ -108,6 +109,7 @@ CFLAGS_eal_pci_vfio.o := -D_GNU_SOURCE CFLAGS_eal_common_whitelist.o := -D_GNU_SOURCE CFLAGS_eal_common_options.o := -D_GNU_SOURCE CFLAGS_eal_common_thread.o := -D_GNU_SOURCE +CFLAGS_eal_common_lcore.o := -D_GNU_SOURCE # workaround for a gcc bug with noreturn attribute # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 diff --git a/lib/librte_eal/linuxapp/eal/eal_lcore.c b/lib/librte_eal/linuxapp/eal/eal_lcore.c index ef8c433..9af6d38 100644 --- a/lib/librte_eal/linuxapp/eal/eal_lcore.c +++ b/lib/librte_eal/linuxapp/eal/eal_lcore.c @@ -52,8 +52,8 @@ #define PHYS_PKG_FILE "topology/physical_package_id" /* Check if a cpu is present by the presence of the cpu information for it */ -static int -cpu_detected(unsigned lcore_id) +int +eal_cpu_detected(unsigned lcore_id) { char path[PATH_MAX]; int len = snprintf(path, sizeof(path), SYS_CPU_DIR @@ -122,8 +122,8 @@ err: } /* Get the cpu core id value from the /sys/.../cpuX core_id value */ -static unsigned -cpu_core_id(unsigned lcore_id) +unsigned +eal_cpu_core_id(unsigned lcore_id) { char path[PATH_MAX]; unsigned long id; @@ -140,61 +140,3 @@ err: "for lcore %u - assuming core 0\n", SYS_CPU_DIR, lcore_id); return 0; } - -/* - * Parse /sys/devices/system/cpu to get the number of physical and logical - * processors on the machine. The function will fill the cpu_info - * structure. - */ -int -rte_eal_cpu_init(void) -{ - /* pointer to global configuration */ - struct rte_config *config = rte_eal_get_configuration(); - unsigned lcore_id; - unsigned count = 0; - - /* - * Parse the maximum set of logical cores, detect the subset of running - * ones and enable them by default. - */ - for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) { - /* init cpuset for per lcore config */ - CPU_ZERO(&lcore_config[lcore_id].cpuset); - - /* in 1:1 mapping, record related cpu detected state */ - lcore_config[lcore_id].detected = cpu_detected(lcore_id); - if (lcore_config[lcore_id].detected == 0) { - config->lcore_role[lcore_id] = ROLE_OFF; - continue; - } - - /* By default, lcore 1:1 map to cpu id */ - CPU_SET(lcore_id, &lcore_config[lcore_id].cpuset); - - /* By default, each detected core is enabled */ - config->lcore_role[lcore_id] = ROLE_RTE; - lcore_config[lcore_id].core_id = cpu_core_id(lcore_id); - lcore_config[lcore_id].socket_id = eal_cpu_socket_id(lcore_id); - if (lcore_config[lcore_id].socket_id >= RTE_MAX_NUMA_NODES) -#ifdef RTE_EAL_ALLOW_INV_SOCKET_ID - lcore_config[lcore_id].socket_id = 0; -#else - rte_panic("Socket ID (%u) is greater than " - "RTE_MAX_NUMA_NODES (%d)\n", - lcore_config[lcore_id].socket_id, RTE_MAX_NUMA_NODES); -#endif - RTE_LOG(DEBUG, EAL, "Detected lcore %u as core %u on socket %u\n", - lcore_id, - lcore_config[lcore_id].core_id, - lcore_config[lcore_id].socket_id); - count ++; - } - /* Set the count of enabled logical cores of the EAL configuration */ - config->lcore_count = count; - RTE_LOG(DEBUG, EAL, "Support maximum %u logical core(s) by configuration.\n", - RTE_MAX_LCORE); - RTE_LOG(DEBUG, EAL, "Detected %u lcore(s)\n", config->lcore_count); - - return 0; -} -- 1.9.1 ^ permalink raw reply [flat|nested] 27+ messages in thread
* [dpdk-dev] [PATCH v7 4/6] Move common functions in eal_timer.c 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 2/6] Move common functions in eal.c Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 3/6] Move common functions in eal_lcore.c Ravi Kerur @ 2015-04-23 21:35 ` Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 5/6] Move common functions in eal_memory.c Ravi Kerur ` (2 subsequent siblings) 5 siblings, 0 replies; 27+ messages in thread From: Ravi Kerur @ 2015-04-23 21:35 UTC (permalink / raw) To: dev Changes in v7 None Changes in v6 Added new line between Copyright and header file inclusion in eal_common_timer.c. Changes in v5 Rebase to latest code. Changes in v4 Removed extern declaration of eal_tsc_resolution_hz, instead provided _set_ API. Make set_tsc_freq_from_clock as wrapper function for BSD. Changes in v3 Changed subject to be more explicit on file name inclusion. Changes in v2 Use common function name set_tsc_freq_from_sysctl for BSD and Linux. Update comments about its actuality in function declaration. Changes in v1 Move common functions in eal_timer.c to librte_eal/common/ eal_common_timer.c file. Following functions are moved to eal_common_timer.c file void rte_delay_us(unsigned us); uint64_t rte_get_tsc_hz(void); static void set_tsc_freq_fallback(void); void set_tsc_freq(void); Makefile changes to reflect new file added. Fix checkpatch warnings and errors. Signed-off-by: Ravi Kerur <rkerur@gmail.com> --- lib/librte_eal/bsdapp/eal/Makefile | 1 + lib/librte_eal/bsdapp/eal/eal_timer.c | 52 +++------------- lib/librte_eal/common/eal_common_timer.c | 102 +++++++++++++++++++++++++++++++ lib/librte_eal/common/eal_private.h | 24 ++++++++ lib/librte_eal/linuxapp/eal/Makefile | 1 + lib/librte_eal/linuxapp/eal/eal_timer.c | 55 ++--------------- 6 files changed, 140 insertions(+), 95 deletions(-) create mode 100644 lib/librte_eal/common/eal_common_timer.c diff --git a/lib/librte_eal/bsdapp/eal/Makefile b/lib/librte_eal/bsdapp/eal/Makefile index 1da0410..d8e6ced 100644 --- a/lib/librte_eal/bsdapp/eal/Makefile +++ b/lib/librte_eal/bsdapp/eal/Makefile @@ -83,6 +83,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_mem_cfg.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_proc_type.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_app_usage.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_lcore.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_BSDAPP) += eal_common_timer.c CFLAGS_eal.o := -D_GNU_SOURCE #CFLAGS_eal_thread.o := -D_GNU_SOURCE diff --git a/lib/librte_eal/bsdapp/eal/eal_timer.c b/lib/librte_eal/bsdapp/eal/eal_timer.c index 7147abe..ee7a5ac 100644 --- a/lib/librte_eal/bsdapp/eal/eal_timer.c +++ b/lib/librte_eal/bsdapp/eal/eal_timer.c @@ -55,29 +55,12 @@ enum timer_source eal_timer_source = EAL_TIMER_TSC; -/* The frequency of the RDTSC timer resolution */ -static uint64_t eal_tsc_resolution_hz = 0; - -void -rte_delay_us(unsigned us) -{ - const uint64_t start = rte_get_timer_cycles(); - const uint64_t ticks = (uint64_t)us * rte_get_timer_hz() / 1E6; - while ((rte_get_timer_cycles() - start) < ticks) - rte_pause(); -} - -uint64_t -rte_get_tsc_hz(void) -{ - return eal_tsc_resolution_hz; -} - static int set_tsc_freq_from_sysctl(void) { size_t sz; int tmp; + uint64_t tsc_hz; sz = sizeof(tmp); tmp = 0; @@ -94,42 +77,23 @@ set_tsc_freq_from_sysctl(void) else if (tmp != 1) RTE_LOG(WARNING, EAL, "TSC is not invariant\n"); - sz = sizeof(eal_tsc_resolution_hz); - if (sysctlbyname("machdep.tsc_freq", &eal_tsc_resolution_hz, &sz, NULL, 0)) { + sz = sizeof(tsc_hz); + if (sysctlbyname("machdep.tsc_freq", &tsc_hz, &sz, NULL, 0)) { RTE_LOG(WARNING, EAL, "%s\n", strerror(errno)); return -1; } + rte_set_tsc_hz(tsc_hz); return 0; } -static void -set_tsc_freq_fallback(void) -{ - RTE_LOG(WARNING, EAL, "WARNING: clock_gettime cannot use " - "CLOCK_MONOTONIC_RAW and HPET is not available" - " - clock timings may be less accurate.\n"); - /* assume that the sleep(1) will sleep for 1 second */ - uint64_t start = rte_rdtsc(); - sleep(1); - eal_tsc_resolution_hz = rte_rdtsc() - start; -} - /* - * This function measures the TSC frequency. It uses a variety of approaches. - * - * 1. Read the TSC frequency value provided by the kernel - * 2. If above does not work, just sleep for 1 second and tune off that, - * printing a warning about inaccuracy of timing + * Wrapper function to get TSC frequency from sysctl. */ -static void -set_tsc_freq(void) +int +set_tsc_freq_from_clock(void) { - if (set_tsc_freq_from_sysctl() < 0) - set_tsc_freq_fallback(); - - RTE_LOG(INFO, EAL, "TSC frequency is ~%"PRIu64" KHz\n", - eal_tsc_resolution_hz/1000); + return set_tsc_freq_from_sysctl(); } int diff --git a/lib/librte_eal/common/eal_common_timer.c b/lib/librte_eal/common/eal_common_timer.c new file mode 100644 index 0000000..5fddd6e --- /dev/null +++ b/lib/librte_eal/common/eal_common_timer.c @@ -0,0 +1,102 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2010-2014 Intel Corporation. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include <string.h> +#include <stdio.h> +#include <unistd.h> +#include <inttypes.h> +#include <sys/types.h> +#include <sys/sysctl.h> +#include <errno.h> + +#include <rte_common.h> +#include <rte_log.h> +#include <rte_cycles.h> + +#include "eal_private.h" +#include "eal_internal_cfg.h" + +/* The frequency of the RDTSC timer resolution */ +uint64_t eal_tsc_resolution_hz = 0; + +void +rte_delay_us(unsigned us) +{ + const uint64_t start = rte_get_timer_cycles(); + const uint64_t ticks = (uint64_t)us * rte_get_timer_hz() / 1E6; + + while ((rte_get_timer_cycles() - start) < ticks) + rte_pause(); +} + +uint64_t +rte_get_tsc_hz(void) +{ + return eal_tsc_resolution_hz; +} + +void +rte_set_tsc_hz(uint64_t tsc_hz) +{ + eal_tsc_resolution_hz = tsc_hz; +} + +static void +set_tsc_freq_fallback(void) +{ + RTE_LOG(WARNING, EAL, "WARNING: clock_gettime cannot use " + "CLOCK_MONOTONIC_RAW and HPET is not available" + " - clock timings may be less accurate.\n"); + /* assume that the sleep(1) will sleep for 1 second */ + uint64_t start = rte_rdtsc(); + + sleep(1); + eal_tsc_resolution_hz = rte_rdtsc() - start; +} + +/* + * This function measures the TSC frequency. It uses a variety of approaches. + * + * 1. Read the TSC frequency value provided by the kernel + * 2. If above does not work, just sleep for 1 second and tune off that, + * printing a warning about inaccuracy of timing + */ +void +set_tsc_freq(void) +{ + if (set_tsc_freq_from_clock() < 0) + set_tsc_freq_fallback(); + + RTE_LOG(INFO, EAL, "TSC frequency is ~%"PRIu64" KHz\n", + eal_tsc_resolution_hz/1000); +} diff --git a/lib/librte_eal/common/eal_private.h b/lib/librte_eal/common/eal_private.h index 76e47a9..8de4ebb 100644 --- a/lib/librte_eal/common/eal_private.h +++ b/lib/librte_eal/common/eal_private.h @@ -324,4 +324,28 @@ unsigned eal_cpu_core_id(unsigned lcore_id); */ int eal_cpu_detected(unsigned lcore_id); +/** + * This function measures TSC frequency + * + * This function is private to the EAL. + */ +void set_tsc_freq(void); + +/** + * This function sets TSC frequency from sysctl + * for BSD and from clock for Linux. + * Is a wrapper function for BSD which will + * internally call set_tsc_freq_from_sysctl. + * + * This function is private to the EAL. + */ +int set_tsc_freq_from_clock(void); + +/** + * This function sets TSC frequency + * + * This function is private to the EAL. + */ +void rte_set_tsc_hz(uint64_t tsc_hz); + #endif /* _EAL_PRIVATE_H_ */ diff --git a/lib/librte_eal/linuxapp/eal/Makefile b/lib/librte_eal/linuxapp/eal/Makefile index 7259bd7..a1c277e 100644 --- a/lib/librte_eal/linuxapp/eal/Makefile +++ b/lib/librte_eal/linuxapp/eal/Makefile @@ -95,6 +95,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_lcore.c +SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_timer.c CFLAGS_eal.o := -D_GNU_SOURCE CFLAGS_eal_interrupts.o := -D_GNU_SOURCE diff --git a/lib/librte_eal/linuxapp/eal/eal_timer.c b/lib/librte_eal/linuxapp/eal/eal_timer.c index 169c6e1..b7484f3 100644 --- a/lib/librte_eal/linuxapp/eal/eal_timer.c +++ b/lib/librte_eal/linuxapp/eal/eal_timer.c @@ -57,9 +57,6 @@ enum timer_source eal_timer_source = EAL_TIMER_HPET; -/* The frequency of the RDTSC timer resolution */ -static uint64_t eal_tsc_resolution_hz = 0; - #ifdef RTE_LIBEAL_USE_HPET #define DEV_HPET "/dev/hpet" @@ -160,23 +157,6 @@ rte_get_hpet_cycles(void) #endif - -void -rte_delay_us(unsigned us) -{ - const uint64_t start = rte_get_timer_cycles(); - const uint64_t ticks = (uint64_t)us * rte_get_timer_hz() / 1E6; - while ((rte_get_timer_cycles() - start) < ticks) - rte_pause(); -} - -uint64_t -rte_get_tsc_hz(void) -{ - return eal_tsc_resolution_hz; -} - - #ifdef RTE_LIBEAL_USE_HPET /* * Open and mmap /dev/hpet (high precision event timer) that will @@ -275,7 +255,7 @@ check_tsc_flags(void) fclose(stream); } -static int +int set_tsc_freq_from_clock(void) { #ifdef CLOCK_MONOTONIC_RAW @@ -284,6 +264,7 @@ set_tsc_freq_from_clock(void) struct timespec sleeptime = {.tv_nsec = 5E8 }; /* 1/2 second */ struct timespec t_start, t_end; + uint64_t tsc_hz; if (clock_gettime(CLOCK_MONOTONIC_RAW, &t_start) == 0) { uint64_t ns, end, start = rte_rdtsc(); @@ -294,42 +275,14 @@ set_tsc_freq_from_clock(void) ns += (t_end.tv_nsec - t_start.tv_nsec); double secs = (double)ns/NS_PER_SEC; - eal_tsc_resolution_hz = (uint64_t)((end - start)/secs); + tsc_hz = (uint64_t)((end - start)/secs); + rte_set_tsc_hz(tsc_hz); return 0; } #endif return -1; } -static void -set_tsc_freq_fallback(void) -{ - RTE_LOG(WARNING, EAL, "WARNING: clock_gettime cannot use " - "CLOCK_MONOTONIC_RAW and HPET is not available" - " - clock timings may be less accurate.\n"); - /* assume that the sleep(1) will sleep for 1 second */ - uint64_t start = rte_rdtsc(); - sleep(1); - eal_tsc_resolution_hz = rte_rdtsc() - start; -} -/* - * This function measures the TSC frequency. It uses a variety of approaches. - * - * 1. If kernel provides CLOCK_MONOTONIC_RAW we use that to tune the TSC value - * 2. If kernel does not provide that, and we have HPET support, tune using HPET - * 3. Lastly, if neither of the above can be used, just sleep for 1 second and - * tune off that, printing a warning about inaccuracy of timing - */ -static void -set_tsc_freq(void) -{ - if (set_tsc_freq_from_clock() < 0) - set_tsc_freq_fallback(); - - RTE_LOG(INFO, EAL, "TSC frequency is ~%"PRIu64" KHz\n", - eal_tsc_resolution_hz/1000); -} - int rte_eal_timer_init(void) { -- 1.9.1 ^ permalink raw reply [flat|nested] 27+ messages in thread
* [dpdk-dev] [PATCH v7 5/6] Move common functions in eal_memory.c 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Ravi Kerur ` (2 preceding siblings ...) 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 4/6] Move common functions in eal_timer.c Ravi Kerur @ 2015-04-23 21:35 ` Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 6/6] Move common functions in eal_pci.c Ravi Kerur 2015-04-24 13:51 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Neil Horman 5 siblings, 0 replies; 27+ messages in thread From: Ravi Kerur @ 2015-04-23 21:35 UTC (permalink / raw) To: dev Changes in v7 None Changes in v6 Removed unnecessary comments in function declaration. Changes in v5 Rebase to latest code. Changes in v4 Make rte_eal_hugepage_init and rte_eal_hugepage_attach as wrapper functions for BSD. Changes in v3 Changed subject to be more explicit on file name inclusion. Changes in v2 Use common function names rte_eal_hugepage_init and rte_eal_hugepage_attach for BSD and Linux. Update comments about its actuality in function declaration. Changes in v1 Move common functions in eal_memory.c to librte_eal/common/ eal_common_memory.c file. Following functions are moved to eal_common_memory.c file static int rte_eal_memdevice_init(void); int rte_eal_memory_init(void); Fix checkpatch warnings and errors. Signed-off-by: Ravi Kerur <rkerur@gmail.com> --- lib/librte_eal/bsdapp/eal/eal_memory.c | 47 +++++++++++-------------------- lib/librte_eal/common/eal_common_memory.c | 38 +++++++++++++++++++++++-- lib/librte_eal/common/eal_private.h | 20 +++++++++++-- lib/librte_eal/linuxapp/eal/eal_memory.c | 36 ++--------------------- 4 files changed, 72 insertions(+), 69 deletions(-) diff --git a/lib/librte_eal/bsdapp/eal/eal_memory.c b/lib/librte_eal/bsdapp/eal/eal_memory.c index 33ebd0f..77c27b3 100644 --- a/lib/librte_eal/bsdapp/eal/eal_memory.c +++ b/lib/librte_eal/bsdapp/eal/eal_memory.c @@ -59,7 +59,7 @@ rte_mem_virt2phy(const void *virtaddr) return RTE_BAD_PHYS_ADDR; } -static int +static inline int rte_eal_contigmem_init(void) { struct rte_mem_config *mcfg; @@ -131,7 +131,16 @@ rte_eal_contigmem_init(void) return 0; } -static int +/* + * Wrapper function to initialize contigmem. + */ +int +rte_eal_hugepage_init(void) +{ + return rte_eal_contigmem_init(); +} + +static inline int rte_eal_contigmem_attach(void) { const struct hugepage_info *hpi; @@ -192,35 +201,11 @@ error: return -1; } - -static int -rte_eal_memdevice_init(void) -{ - struct rte_config *config; - - if (rte_eal_process_type() == RTE_PROC_SECONDARY) - return 0; - - config = rte_eal_get_configuration(); - config->mem_config->nchannel = internal_config.force_nchannel; - config->mem_config->nrank = internal_config.force_nrank; - - return 0; -} - -/* init memory subsystem */ +/* + * Wrapper function to attach contigmem. + */ int -rte_eal_memory_init(void) +rte_eal_hugepage_attach(void) { - RTE_LOG(INFO, EAL, "Setting up physically contiguous memory...\n"); - const int retval = rte_eal_process_type() == RTE_PROC_PRIMARY ? - rte_eal_contigmem_init() : - rte_eal_contigmem_attach(); - if (retval < 0) - return -1; - - if (internal_config.no_shconf == 0 && rte_eal_memdevice_init() < 0) - return -1; - - return 0; + return rte_eal_contigmem_attach(); } diff --git a/lib/librte_eal/common/eal_common_memory.c b/lib/librte_eal/common/eal_common_memory.c index 9a07b1e..10ff0bc 100644 --- a/lib/librte_eal/common/eal_common_memory.c +++ b/lib/librte_eal/common/eal_common_memory.c @@ -45,6 +45,7 @@ #include <rte_log.h> #include "eal_private.h" +#include "eal_internal_cfg.h" /* * Return a pointer to a read-only table of struct rte_physmem_desc @@ -69,7 +70,7 @@ rte_eal_get_physmem_size(void) /* get pointer to global configuration */ mcfg = rte_eal_get_configuration()->mem_config; - for (i=0; i<RTE_MAX_MEMSEG; i++) { + for (i = 0; i < RTE_MAX_MEMSEG; i++) { if (mcfg->memseg[i].addr == NULL) break; @@ -89,7 +90,7 @@ rte_dump_physmem_layout(FILE *f) /* get pointer to global configuration */ mcfg = rte_eal_get_configuration()->mem_config; - for (i=0; i<RTE_MAX_MEMSEG; i++) { + for (i = 0; i < RTE_MAX_MEMSEG; i++) { if (mcfg->memseg[i].addr == NULL) break; @@ -118,3 +119,36 @@ unsigned rte_memory_get_nrank(void) { return rte_eal_get_configuration()->mem_config->nrank; } + +static int +rte_eal_memdevice_init(void) +{ + struct rte_config *config; + + if (rte_eal_process_type() == RTE_PROC_SECONDARY) + return 0; + + config = rte_eal_get_configuration(); + config->mem_config->nchannel = internal_config.force_nchannel; + config->mem_config->nrank = internal_config.force_nrank; + + return 0; +} + +/* init memory subsystem */ +int +rte_eal_memory_init(void) +{ + RTE_LOG(INFO, EAL, "Setting up physically contiguous memory...\n"); + const int retval = rte_eal_process_type() == RTE_PROC_PRIMARY ? + rte_eal_hugepage_init() : + rte_eal_hugepage_attach(); + + if (retval < 0) + return -1; + + if (internal_config.no_shconf == 0 && rte_eal_memdevice_init() < 0) + return -1; + + return 0; +} diff --git a/lib/librte_eal/common/eal_private.h b/lib/librte_eal/common/eal_private.h index 8de4ebb..a51a68e 100644 --- a/lib/librte_eal/common/eal_private.h +++ b/lib/librte_eal/common/eal_private.h @@ -334,8 +334,6 @@ void set_tsc_freq(void); /** * This function sets TSC frequency from sysctl * for BSD and from clock for Linux. - * Is a wrapper function for BSD which will - * internally call set_tsc_freq_from_sysctl. * * This function is private to the EAL. */ @@ -348,4 +346,22 @@ int set_tsc_freq_from_clock(void); */ void rte_set_tsc_hz(uint64_t tsc_hz); +/** + * This function prepares physical memory mapping + * i.e. hugepages on Linux and + * contigmem on BSD. + * + * This function is private to the EAL. + */ +int rte_eal_hugepage_init(void); + +/** + * This function creates memory mapping in secondary + * i.e. hugepages on Linux and + * contigmem on BSD. + * + * This function is private to the EAL. + */ +int rte_eal_hugepage_attach(void); + #endif /* _EAL_PRIVATE_H_ */ diff --git a/lib/librte_eal/linuxapp/eal/eal_memory.c b/lib/librte_eal/linuxapp/eal/eal_memory.c index 5f9f92e..c83a2cb 100644 --- a/lib/librte_eal/linuxapp/eal/eal_memory.c +++ b/lib/librte_eal/linuxapp/eal/eal_memory.c @@ -1030,7 +1030,7 @@ calc_num_pages_per_socket(uint64_t * memory, * 6. unmap the first mapping * 7. fill memsegs in configuration with contiguous zones */ -static int +int rte_eal_hugepage_init(void) { struct rte_mem_config *mcfg; @@ -1368,7 +1368,7 @@ getFileSize(int fd) * configuration and finds the hugepages which form that segment, mapping them * in order to form a contiguous block in the virtual memory space */ -static int +int rte_eal_hugepage_attach(void) { const struct rte_mem_config *mcfg = rte_eal_get_configuration()->mem_config; @@ -1529,35 +1529,3 @@ error: return -1; } -static int -rte_eal_memdevice_init(void) -{ - struct rte_config *config; - - if (rte_eal_process_type() == RTE_PROC_SECONDARY) - return 0; - - config = rte_eal_get_configuration(); - config->mem_config->nchannel = internal_config.force_nchannel; - config->mem_config->nrank = internal_config.force_nrank; - - return 0; -} - - -/* init memory subsystem */ -int -rte_eal_memory_init(void) -{ - RTE_LOG(INFO, EAL, "Setting up memory...\n"); - const int retval = rte_eal_process_type() == RTE_PROC_PRIMARY ? - rte_eal_hugepage_init() : - rte_eal_hugepage_attach(); - if (retval < 0) - return -1; - - if (internal_config.no_shconf == 0 && rte_eal_memdevice_init() < 0) - return -1; - - return 0; -} -- 1.9.1 ^ permalink raw reply [flat|nested] 27+ messages in thread
* [dpdk-dev] [PATCH v7 6/6] Move common functions in eal_pci.c 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Ravi Kerur ` (3 preceding siblings ...) 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 5/6] Move common functions in eal_memory.c Ravi Kerur @ 2015-04-23 21:35 ` Ravi Kerur 2015-04-24 13:51 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Neil Horman 5 siblings, 0 replies; 27+ messages in thread From: Ravi Kerur @ 2015-04-23 21:35 UTC (permalink / raw) To: dev Changes in v7 Rebase to latest code. Changes in v6 Split changes due to complexity. v6 includes moving rte_eal_pci_probe_one_driver function and its associated utility functions only. Changes in v5 Rebase to latest code. Removed RTE_EXEC_ENV_BSDAPP from earlier changes. Changes in v4 Move common functions in eal_pci.c to librte_eal/common/ eal_common_pci.c file. Following functions are moved to eal_common_pci.c file. void *pci_map_resource(void *requested_addr, const int vfio_fd, const char *devname, off_t offset, size_t size); int pci_addr_comparison(struct rte_pci_addr *addr, struct rte_pci_addr *addr2); int rte_eal_pci_probe_one_driver(struct rte_pci_driver *dr, struct rte_pci_device *dev); Use RTE_EXEC_ENV_BSDAPP to differentiate minor differences in common function. Fix checkpatch warnings and errors. Changes in v3 N/A Changes in v2 N/A Changes in v1 N/A Signed-off-by: Ravi Kerur <rkerur@gmail.com> --- lib/librte_eal/bsdapp/eal/eal_pci.c | 72 +++++--------------------------- lib/librte_eal/common/eal_common_pci.c | 72 ++++++++++++++++++++++++++++++++ lib/librte_eal/common/eal_private.h | 39 +++++++++++++----- lib/librte_eal/linuxapp/eal/eal_pci.c | 75 +--------------------------------- 4 files changed, 113 insertions(+), 145 deletions(-) diff --git a/lib/librte_eal/bsdapp/eal/eal_pci.c b/lib/librte_eal/bsdapp/eal/eal_pci.c index 30f0232..f21b5b6 100644 --- a/lib/librte_eal/bsdapp/eal/eal_pci.c +++ b/lib/librte_eal/bsdapp/eal/eal_pci.c @@ -111,7 +111,7 @@ static struct rte_tailq_elem rte_uio_tailq = { EAL_REGISTER_TAILQ(rte_uio_tailq) /* unbind kernel driver for this device */ -static int +int pci_unbind_kernel_driver(struct rte_pci_device *dev __rte_unused) { RTE_LOG(ERR, EAL, "RTE_PCI_DRV_FORCE_UNBIND flag is not implemented " @@ -274,6 +274,13 @@ pci_uio_map_resource(struct rte_pci_device *dev) return (0); } +/* map the PCI resource of a PCI device in virtual memory */ +int +pci_map_device(struct rte_pci_device *dev) +{ + return pci_uio_map_resource(dev); +} + /* Scan one pci sysfs entry, and fill the devices list from it. */ static int pci_scan_one(int dev_pci_fd, struct pci_conf *conf) @@ -426,68 +433,11 @@ error: } /* - * If vendor/device ID match, call the devinit() function of the - * driver. + * This function is a no-op in BSD. */ -int -rte_eal_pci_probe_one_driver(struct rte_pci_driver *dr, struct rte_pci_device *dev) +void +pci_config_space_set(struct rte_pci_device *dev __rte_unused) { - struct rte_pci_id *id_table; - int ret; - - for (id_table = dr->id_table ; id_table->vendor_id != 0; id_table++) { - - /* check if device's identifiers match the driver's ones */ - if (id_table->vendor_id != dev->id.vendor_id && - id_table->vendor_id != PCI_ANY_ID) - continue; - if (id_table->device_id != dev->id.device_id && - id_table->device_id != PCI_ANY_ID) - continue; - if (id_table->subsystem_vendor_id != dev->id.subsystem_vendor_id && - id_table->subsystem_vendor_id != PCI_ANY_ID) - continue; - if (id_table->subsystem_device_id != dev->id.subsystem_device_id && - id_table->subsystem_device_id != PCI_ANY_ID) - continue; - - struct rte_pci_addr *loc = &dev->addr; - - RTE_LOG(DEBUG, EAL, "PCI device "PCI_PRI_FMT" on NUMA socket %i\n", - loc->domain, loc->bus, loc->devid, loc->function, - dev->numa_node); - - RTE_LOG(DEBUG, EAL, " probe driver: %x:%x %s\n", dev->id.vendor_id, - dev->id.device_id, dr->name); - - /* no initialization when blacklisted, return without error */ - if (dev->devargs != NULL && - dev->devargs->type == RTE_DEVTYPE_BLACKLISTED_PCI) { - - RTE_LOG(DEBUG, EAL, " Device is blacklisted, not initializing\n"); - return 0; - } - - if (dr->drv_flags & RTE_PCI_DRV_NEED_MAPPING) { - /* map resources for devices that use igb_uio */ - ret = pci_uio_map_resource(dev); - if (ret != 0) - return ret; - } else if (dr->drv_flags & RTE_PCI_DRV_FORCE_UNBIND && - rte_eal_process_type() == RTE_PROC_PRIMARY) { - /* unbind current driver */ - if (pci_unbind_kernel_driver(dev) < 0) - return -1; - } - - /* reference driver structure */ - dev->driver = dr; - - /* call the driver devinit() function */ - return dr->devinit(dr, dev); - } - /* return positive value if driver is not found */ - return 1; } /* Init the PCI EAL subsystem */ diff --git a/lib/librte_eal/common/eal_common_pci.c b/lib/librte_eal/common/eal_common_pci.c index 4229aaf..279063f 100644 --- a/lib/librte_eal/common/eal_common_pci.c +++ b/lib/librte_eal/common/eal_common_pci.c @@ -99,6 +99,78 @@ static struct rte_devargs *pci_devargs_lookup(struct rte_pci_device *dev) } /* + * If vendor/device ID match, call the devinit() function of the + * driver. + */ +static int +rte_eal_pci_probe_one_driver(struct rte_pci_driver *dr, + struct rte_pci_device *dev) +{ + int ret; + const struct rte_pci_id *id_table; + + for (id_table = dr->id_table ; id_table->vendor_id != 0; id_table++) { + + /* check if device's identifiers match the driver's ones */ + if (id_table->vendor_id != dev->id.vendor_id && + id_table->vendor_id != PCI_ANY_ID) + continue; + if (id_table->device_id != dev->id.device_id && + id_table->device_id != PCI_ANY_ID) + continue; + if (id_table->subsystem_vendor_id != dev->id.subsystem_vendor_id && + id_table->subsystem_vendor_id != PCI_ANY_ID) + continue; + if (id_table->subsystem_device_id != dev->id.subsystem_device_id && + id_table->subsystem_device_id != PCI_ANY_ID) + continue; + + struct rte_pci_addr *loc = &dev->addr; + + RTE_LOG(DEBUG, EAL, "PCI device "PCI_PRI_FMT" on NUMA socket %i\n", + loc->domain, loc->bus, loc->devid, loc->function, + dev->numa_node); + + RTE_LOG(DEBUG, EAL, " probe driver: %x:%x %s\n", dev->id.vendor_id, + dev->id.device_id, dr->name); + + /* no initialization when blacklisted, return without error */ + if (dev->devargs != NULL && + dev->devargs->type == RTE_DEVTYPE_BLACKLISTED_PCI) { + RTE_LOG(DEBUG, EAL, " Device is blacklisted, not initializing\n"); + return 1; + } + + if (dr->drv_flags & RTE_PCI_DRV_NEED_MAPPING) { +#ifdef RTE_PCI_CONFIG + /* + * Set PCIe config space for high performance. + * Return value can be ignored. + */ + pci_config_space_set(dev); +#endif + /* map resources for devices that use igb_uio */ + ret = pci_map_device(dev); + if (ret != 0) + return ret; + } else if (dr->drv_flags & RTE_PCI_DRV_FORCE_UNBIND && + rte_eal_process_type() == RTE_PROC_PRIMARY) { + /* unbind current driver */ + if (pci_unbind_kernel_driver(dev) < 0) + return -1; + } + + /* reference driver structure */ + dev->driver = dr; + + /* call the driver devinit() function */ + return dr->devinit(dr, dev); + } + /* return positive value if driver is not found */ + return 1; +} + +/* * If vendor/device ID match, call the devinit() function of all * registered driver for the given device. Return -1 if initialization * failed, return 1 if no driver is found for this device. diff --git a/lib/librte_eal/common/eal_private.h b/lib/librte_eal/common/eal_private.h index a51a68e..6a9d13e 100644 --- a/lib/librte_eal/common/eal_private.h +++ b/lib/librte_eal/common/eal_private.h @@ -156,17 +156,6 @@ struct rte_pci_driver; struct rte_pci_device; /** - * Mmap memory for single PCI device - * - * This function is private to EAL. - * - * @return - * 0 on success, negative on error - */ -int rte_eal_pci_probe_one_driver(struct rte_pci_driver *dr, - struct rte_pci_device *dev); - -/** * Munmap memory for single PCI device * * This function is private to EAL. @@ -364,4 +353,32 @@ int rte_eal_hugepage_init(void); */ int rte_eal_hugepage_attach(void); +/** + * This function unbinds kernel driver for this device + * + * This function is private to the EAL. + */ +int +pci_unbind_kernel_driver(struct rte_pci_device *dev); + +/** + * This function maps resources for devices + * that use igb_uio on Linux and it's a + * wrapper for pci_uio_map_resource on BSD. + * + * This function is private to the EAL. + */ +int +pci_map_device(struct rte_pci_device *dev); + +/** + * This function sets PCIe config space for + * high performance. + * It's a NO-OP on BSD. + * + * This function is private to the EAL. + */ +void +pci_config_space_set(struct rte_pci_device *dev); + #endif /* _EAL_PRIVATE_H_ */ diff --git a/lib/librte_eal/linuxapp/eal/eal_pci.c b/lib/librte_eal/linuxapp/eal/eal_pci.c index d2adc66..7b6378d 100644 --- a/lib/librte_eal/linuxapp/eal/eal_pci.c +++ b/lib/librte_eal/linuxapp/eal/eal_pci.c @@ -57,7 +57,7 @@ */ /* unbind kernel driver for this device */ -static int +int pci_unbind_kernel_driver(struct rte_pci_device *dev) { int n; @@ -563,7 +563,7 @@ pci_config_space_set(struct rte_pci_device *dev) } #endif -static int +int pci_map_device(struct rte_pci_device *dev) { int ret = -1; @@ -616,77 +616,6 @@ pci_unmap_device(struct rte_pci_device *dev) } #endif /* RTE_LIBRTE_EAL_HOTPLUG */ -/* - * If vendor/device ID match, call the devinit() function of the - * driver. - */ -int -rte_eal_pci_probe_one_driver(struct rte_pci_driver *dr, struct rte_pci_device *dev) -{ - int ret; - const struct rte_pci_id *id_table; - - for (id_table = dr->id_table; id_table->vendor_id != 0; id_table++) { - - /* check if device's identifiers match the driver's ones */ - if (id_table->vendor_id != dev->id.vendor_id && - id_table->vendor_id != PCI_ANY_ID) - continue; - if (id_table->device_id != dev->id.device_id && - id_table->device_id != PCI_ANY_ID) - continue; - if (id_table->subsystem_vendor_id != dev->id.subsystem_vendor_id && - id_table->subsystem_vendor_id != PCI_ANY_ID) - continue; - if (id_table->subsystem_device_id != dev->id.subsystem_device_id && - id_table->subsystem_device_id != PCI_ANY_ID) - continue; - - struct rte_pci_addr *loc = &dev->addr; - - RTE_LOG(DEBUG, EAL, "PCI device "PCI_PRI_FMT" on NUMA socket %i\n", - loc->domain, loc->bus, loc->devid, loc->function, - dev->numa_node); - - RTE_LOG(DEBUG, EAL, " probe driver: %x:%x %s\n", dev->id.vendor_id, - dev->id.device_id, dr->name); - - /* no initialization when blacklisted, return without error */ - if (dev->devargs != NULL && - dev->devargs->type == RTE_DEVTYPE_BLACKLISTED_PCI) { - RTE_LOG(DEBUG, EAL, " Device is blacklisted, not initializing\n"); - return 1; - } - - if (dr->drv_flags & RTE_PCI_DRV_NEED_MAPPING) { -#ifdef RTE_PCI_CONFIG - /* - * Set PCIe config space for high performance. - * Return value can be ignored. - */ - pci_config_space_set(dev); -#endif - /* map resources for devices that use igb_uio */ - ret = pci_map_device(dev); - if (ret != 0) - return ret; - } else if (dr->drv_flags & RTE_PCI_DRV_FORCE_UNBIND && - rte_eal_process_type() == RTE_PROC_PRIMARY) { - /* unbind current driver */ - if (pci_unbind_kernel_driver(dev) < 0) - return -1; - } - - /* reference driver structure */ - dev->driver = dr; - - /* call the driver devinit() function */ - return dr->devinit(dr, dev); - } - /* return positive value if driver is not found */ - return 1; -} - #ifdef RTE_LIBRTE_EAL_HOTPLUG /* * If vendor/device ID match, call the devuninit() function of the -- 1.9.1 ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Ravi Kerur ` (4 preceding siblings ...) 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 6/6] Move common functions in eal_pci.c Ravi Kerur @ 2015-04-24 13:51 ` Neil Horman 2015-04-24 15:14 ` Ravi Kerur 5 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-24 13:51 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > Changes in v7 > Remove _setname_ pthread calls. > Use rte_gettid() API in RTE_LOG to print thread_id. > > Changes in v6 > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > respectively. Plan to use _getname_ in RTE_LOG when available. > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > Changes in v5 > Rebase to latest code. > > Changes in v4 > None > > Changes in v3 > Changed subject to be more explicit on file name inclusion. > > Changes in v2 > None > > Changes in v1 > eal_thread.c has minor differences between Linux and BSD, move > entire file into common directory. > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor differences. > Rename eal_thread.c to eal_common_thread.c > Makefile changes to reflect file move and name change. > Fix checkpatch warnings. > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > --- > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 ------------------------------ > lib/librte_eal/common/eal_common_thread.c | 147 ++++++++++++++++++++++++++++- > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 +----------------------------- > 4 files changed, 148 insertions(+), 305 deletions(-) > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile b/lib/librte_eal/bsdapp/eal/Makefile > index 2357cfa..55971b9 100644 > --- a/lib/librte_eal/bsdapp/eal/Makefile > +++ b/lib/librte_eal/bsdapp/eal/Makefile > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > # workaround for a gcc bug with noreturn attribute > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > -CFLAGS_eal_thread.o += -Wno-return-type > +CFLAGS_eal_common_thread.o += -Wno-return-type > CFLAGS_eal_hpet.o += -Wno-return-type > endif > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c b/lib/librte_eal/bsdapp/eal/eal_thread.c > index 9a03437..5714b8f 100644 > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > @@ -35,163 +35,11 @@ > #include <stdio.h> > #include <stdlib.h> > #include <stdint.h> > -#include <unistd.h> > -#include <sched.h> > -#include <pthread_np.h> > -#include <sys/queue.h> > #include <sys/thr.h> > > -#include <rte_debug.h> > -#include <rte_atomic.h> > -#include <rte_launch.h> > -#include <rte_log.h> > -#include <rte_memory.h> > -#include <rte_memzone.h> > -#include <rte_per_lcore.h> > -#include <rte_eal.h> > -#include <rte_per_lcore.h> > -#include <rte_lcore.h> > - > #include "eal_private.h" > #include "eal_thread.h" > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; NAK, these are exported symbols, you can't remove them without going through the deprecation process. Regards Neil ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-24 13:51 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Neil Horman @ 2015-04-24 15:14 ` Ravi Kerur 2015-04-24 15:22 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-24 15:14 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > > Changes in v7 > > Remove _setname_ pthread calls. > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > > Changes in v6 > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > > respectively. Plan to use _getname_ in RTE_LOG when available. > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > > > Changes in v5 > > Rebase to latest code. > > > > Changes in v4 > > None > > > > Changes in v3 > > Changed subject to be more explicit on file name inclusion. > > > > Changes in v2 > > None > > > > Changes in v1 > > eal_thread.c has minor differences between Linux and BSD, move > > entire file into common directory. > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor differences. > > Rename eal_thread.c to eal_common_thread.c > > Makefile changes to reflect file move and name change. > > Fix checkpatch warnings. > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > --- > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > ------------------------------ > > lib/librte_eal/common/eal_common_thread.c | 147 > ++++++++++++++++++++++++++++- > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > +----------------------------- > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > b/lib/librte_eal/bsdapp/eal/Makefile > > index 2357cfa..55971b9 100644 > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > > # workaround for a gcc bug with noreturn attribute > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > -CFLAGS_eal_thread.o += -Wno-return-type > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > CFLAGS_eal_hpet.o += -Wno-return-type > > endif > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > index 9a03437..5714b8f 100644 > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > @@ -35,163 +35,11 @@ > > #include <stdio.h> > > #include <stdlib.h> > > #include <stdint.h> > > -#include <unistd.h> > > -#include <sched.h> > > -#include <pthread_np.h> > > -#include <sys/queue.h> > > #include <sys/thr.h> > > > > -#include <rte_debug.h> > > -#include <rte_atomic.h> > > -#include <rte_launch.h> > > -#include <rte_log.h> > > -#include <rte_memory.h> > > -#include <rte_memzone.h> > > -#include <rte_per_lcore.h> > > -#include <rte_eal.h> > > -#include <rte_per_lcore.h> > > -#include <rte_lcore.h> > > - > > #include "eal_private.h" > > #include "eal_thread.h" > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > NAK, these are exported symbols, you can't remove them without going > through the > deprecation process. > > They are not removed/deleted, they are moved from eal_thread.c to eal_common_thread.c file since it is common to both Linux and BSD. Thanks Ravi Regards > Neil > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-24 15:14 ` Ravi Kerur @ 2015-04-24 15:22 ` Neil Horman 2015-04-24 16:45 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-24 15:22 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > > > Changes in v7 > > > Remove _setname_ pthread calls. > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > > > > Changes in v6 > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > > > respectively. Plan to use _getname_ in RTE_LOG when available. > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > > > > > Changes in v5 > > > Rebase to latest code. > > > > > > Changes in v4 > > > None > > > > > > Changes in v3 > > > Changed subject to be more explicit on file name inclusion. > > > > > > Changes in v2 > > > None > > > > > > Changes in v1 > > > eal_thread.c has minor differences between Linux and BSD, move > > > entire file into common directory. > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor differences. > > > Rename eal_thread.c to eal_common_thread.c > > > Makefile changes to reflect file move and name change. > > > Fix checkpatch warnings. > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > --- > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > ------------------------------ > > > lib/librte_eal/common/eal_common_thread.c | 147 > > ++++++++++++++++++++++++++++- > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > +----------------------------- > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > b/lib/librte_eal/bsdapp/eal/Makefile > > > index 2357cfa..55971b9 100644 > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > > > # workaround for a gcc bug with noreturn attribute > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > endif > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > index 9a03437..5714b8f 100644 > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > @@ -35,163 +35,11 @@ > > > #include <stdio.h> > > > #include <stdlib.h> > > > #include <stdint.h> > > > -#include <unistd.h> > > > -#include <sched.h> > > > -#include <pthread_np.h> > > > -#include <sys/queue.h> > > > #include <sys/thr.h> > > > > > > -#include <rte_debug.h> > > > -#include <rte_atomic.h> > > > -#include <rte_launch.h> > > > -#include <rte_log.h> > > > -#include <rte_memory.h> > > > -#include <rte_memzone.h> > > > -#include <rte_per_lcore.h> > > > -#include <rte_eal.h> > > > -#include <rte_per_lcore.h> > > > -#include <rte_lcore.h> > > > - > > > #include "eal_private.h" > > > #include "eal_thread.h" > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > > NAK, these are exported symbols, you can't remove them without going > > through the > > deprecation process. > > > > > They are not removed/deleted, they are moved from eal_thread.c to > eal_common_thread.c file since it is common to both Linux and BSD. > Then perhaps you forgot to export the symbol? Its showing up as removed on the ABI checker utility. Neil > Thanks > Ravi > > Regards > > Neil > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-24 15:22 ` Neil Horman @ 2015-04-24 16:45 ` Ravi Kerur 2015-04-24 18:53 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-24 16:45 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman <nhorman@tuxdriver.com> > wrote: > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > > > > Changes in v7 > > > > Remove _setname_ pthread calls. > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > > > > > > Changes in v6 > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > > > > respectively. Plan to use _getname_ in RTE_LOG when available. > > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > > > > > > > Changes in v5 > > > > Rebase to latest code. > > > > > > > > Changes in v4 > > > > None > > > > > > > > Changes in v3 > > > > Changed subject to be more explicit on file name inclusion. > > > > > > > > Changes in v2 > > > > None > > > > > > > > Changes in v1 > > > > eal_thread.c has minor differences between Linux and BSD, move > > > > entire file into common directory. > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor differences. > > > > Rename eal_thread.c to eal_common_thread.c > > > > Makefile changes to reflect file move and name change. > > > > Fix checkpatch warnings. > > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > --- > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > ------------------------------ > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > ++++++++++++++++++++++++++++- > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > +----------------------------- > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > index 2357cfa..55971b9 100644 > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > > > > # workaround for a gcc bug with noreturn attribute > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > endif > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > index 9a03437..5714b8f 100644 > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > @@ -35,163 +35,11 @@ > > > > #include <stdio.h> > > > > #include <stdlib.h> > > > > #include <stdint.h> > > > > -#include <unistd.h> > > > > -#include <sched.h> > > > > -#include <pthread_np.h> > > > > -#include <sys/queue.h> > > > > #include <sys/thr.h> > > > > > > > > -#include <rte_debug.h> > > > > -#include <rte_atomic.h> > > > > -#include <rte_launch.h> > > > > -#include <rte_log.h> > > > > -#include <rte_memory.h> > > > > -#include <rte_memzone.h> > > > > -#include <rte_per_lcore.h> > > > > -#include <rte_eal.h> > > > > -#include <rte_per_lcore.h> > > > > -#include <rte_lcore.h> > > > > - > > > > #include "eal_private.h" > > > > #include "eal_thread.h" > > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > > > NAK, these are exported symbols, you can't remove them without going > > > through the > > > deprecation process. > > > > > > > > They are not removed/deleted, they are moved from eal_thread.c to > > eal_common_thread.c file since it is common to both Linux and BSD. > > > Then perhaps you forgot to export the symbol? Its showing up as removed > on the > ABI checker utility. > > Neil > Can you please show me in the current code where it is being exported? I have only moved definitions to _common_ files, not sure why it should be exported now. I searched in the current code for RTE_DEFINE_PER_LCORE #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * app/test/test_per_lcore.c:static RTE_DEFINE_PER_LCORE(unsigned, test) = 0x12345678; lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, _socket_id) = (unsigned)SOCKET_ID_ANY; lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, _cpuset); lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, _socket_id) = (unsigned)SOCKET_ID_ANY; lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, _cpuset); lib/librte_eal/common/include/rte_per_lcore.h:#define RTE_DEFINE_PER_LCORE(type, name) \ lib/librte_eal/common/include/rte_eal.h: static RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, _rte_errno); lib/librte_eal/common/eal_common_errno.c: static RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > Thanks > > Ravi > > > > Regards > > > Neil > > > > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-24 16:45 ` Ravi Kerur @ 2015-04-24 18:53 ` Neil Horman 2015-04-24 19:21 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-24 18:53 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman <nhorman@tuxdriver.com> > > wrote: > > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > > > > > Changes in v7 > > > > > Remove _setname_ pthread calls. > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > > > > > > > > Changes in v6 > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > > > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > > > > > respectively. Plan to use _getname_ in RTE_LOG when available. > > > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > > > > > > > > > Changes in v5 > > > > > Rebase to latest code. > > > > > > > > > > Changes in v4 > > > > > None > > > > > > > > > > Changes in v3 > > > > > Changed subject to be more explicit on file name inclusion. > > > > > > > > > > Changes in v2 > > > > > None > > > > > > > > > > Changes in v1 > > > > > eal_thread.c has minor differences between Linux and BSD, move > > > > > entire file into common directory. > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor differences. > > > > > Rename eal_thread.c to eal_common_thread.c > > > > > Makefile changes to reflect file move and name change. > > > > > Fix checkpatch warnings. > > > > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > > --- > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > > ------------------------------ > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > > ++++++++++++++++++++++++++++- > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > > +----------------------------- > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > > index 2357cfa..55971b9 100644 > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > > > > > # workaround for a gcc bug with noreturn attribute > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > > endif > > > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > index 9a03437..5714b8f 100644 > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > @@ -35,163 +35,11 @@ > > > > > #include <stdio.h> > > > > > #include <stdlib.h> > > > > > #include <stdint.h> > > > > > -#include <unistd.h> > > > > > -#include <sched.h> > > > > > -#include <pthread_np.h> > > > > > -#include <sys/queue.h> > > > > > #include <sys/thr.h> > > > > > > > > > > -#include <rte_debug.h> > > > > > -#include <rte_atomic.h> > > > > > -#include <rte_launch.h> > > > > > -#include <rte_log.h> > > > > > -#include <rte_memory.h> > > > > > -#include <rte_memzone.h> > > > > > -#include <rte_per_lcore.h> > > > > > -#include <rte_eal.h> > > > > > -#include <rte_per_lcore.h> > > > > > -#include <rte_lcore.h> > > > > > - > > > > > #include "eal_private.h" > > > > > #include "eal_thread.h" > > > > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > > > > NAK, these are exported symbols, you can't remove them without going > > > > through the > > > > deprecation process. > > > > > > > > > > > They are not removed/deleted, they are moved from eal_thread.c to > > > eal_common_thread.c file since it is common to both Linux and BSD. > > > > > Then perhaps you forgot to export the symbol? Its showing up as removed > > on the > > ABI checker utility. > > > > Neil > > > > Can you please show me in the current code where it is being exported? I > have only moved definitions to _common_ files, not sure why it should be > exported now. I searched in the current code for RTE_DEFINE_PER_LCORE > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * > app/test/test_per_lcore.c:static RTE_DEFINE_PER_LCORE(unsigned, test) = > 0x12345678; > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > _lcore_id) = LCORE_ID_ANY; > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > _socket_id) = (unsigned)SOCKET_ID_ANY; > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > _cpuset); > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > _lcore_id) = LCORE_ID_ANY; > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > _socket_id) = (unsigned)SOCKET_ID_ANY; > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > _cpuset); > lib/librte_eal/common/include/rte_per_lcore.h:#define > RTE_DEFINE_PER_LCORE(type, name) \ > lib/librte_eal/common/include/rte_eal.h: static > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > _rte_errno); > lib/librte_eal/common/eal_common_errno.c: static > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > > Thanks > > > Ravi > > > > > > Regards > > > > Neil > > > > > > > > > > Its exported in the version map file: per_lcore__lcore_id; Neil ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-24 18:53 ` Neil Horman @ 2015-04-24 19:21 ` Ravi Kerur 2015-04-24 19:51 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-24 19:21 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman <nhorman@tuxdriver.com> > wrote: > > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman <nhorman@tuxdriver.com> > > > wrote: > > > > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > > > > > > Changes in v7 > > > > > > Remove _setname_ pthread calls. > > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > > > > > > > > > > Changes in v6 > > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > > > > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > > > > > > respectively. Plan to use _getname_ in RTE_LOG when available. > > > > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > > > > > > > > > > > Changes in v5 > > > > > > Rebase to latest code. > > > > > > > > > > > > Changes in v4 > > > > > > None > > > > > > > > > > > > Changes in v3 > > > > > > Changed subject to be more explicit on file name inclusion. > > > > > > > > > > > > Changes in v2 > > > > > > None > > > > > > > > > > > > Changes in v1 > > > > > > eal_thread.c has minor differences between Linux and BSD, move > > > > > > entire file into common directory. > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor differences. > > > > > > Rename eal_thread.c to eal_common_thread.c > > > > > > Makefile changes to reflect file move and name change. > > > > > > Fix checkpatch warnings. > > > > > > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > > > --- > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > > > ------------------------------ > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > > > ++++++++++++++++++++++++++++- > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > > > +----------------------------- > > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > > > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > index 2357cfa..55971b9 100644 > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > > > > > > # workaround for a gcc bug with noreturn attribute > > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > > > endif > > > > > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > index 9a03437..5714b8f 100644 > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > @@ -35,163 +35,11 @@ > > > > > > #include <stdio.h> > > > > > > #include <stdlib.h> > > > > > > #include <stdint.h> > > > > > > -#include <unistd.h> > > > > > > -#include <sched.h> > > > > > > -#include <pthread_np.h> > > > > > > -#include <sys/queue.h> > > > > > > #include <sys/thr.h> > > > > > > > > > > > > -#include <rte_debug.h> > > > > > > -#include <rte_atomic.h> > > > > > > -#include <rte_launch.h> > > > > > > -#include <rte_log.h> > > > > > > -#include <rte_memory.h> > > > > > > -#include <rte_memzone.h> > > > > > > -#include <rte_per_lcore.h> > > > > > > -#include <rte_eal.h> > > > > > > -#include <rte_per_lcore.h> > > > > > > -#include <rte_lcore.h> > > > > > > - > > > > > > #include "eal_private.h" > > > > > > #include "eal_thread.h" > > > > > > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > > > > > NAK, these are exported symbols, you can't remove them without > going > > > > > through the > > > > > deprecation process. > > > > > > > > > > > > > > They are not removed/deleted, they are moved from eal_thread.c to > > > > eal_common_thread.c file since it is common to both Linux and BSD. > > > > > > > Then perhaps you forgot to export the symbol? Its showing up as > removed > > > on the > > > ABI checker utility. > > > > > > Neil > > > > > > > Can you please show me in the current code where it is being exported? I > > have only moved definitions to _common_ files, not sure why it should be > > exported now. I searched in the current code for RTE_DEFINE_PER_LCORE > > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * > > app/test/test_per_lcore.c:static RTE_DEFINE_PER_LCORE(unsigned, test) = > > 0x12345678; > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > _lcore_id) = LCORE_ID_ANY; > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > _cpuset); > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > _lcore_id) = LCORE_ID_ANY; > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > _cpuset); > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > RTE_DEFINE_PER_LCORE(type, name) \ > > lib/librte_eal/common/include/rte_eal.h: static > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > _rte_errno); > > lib/librte_eal/common/eal_common_errno.c: static > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > > > > > Thanks > > > > Ravi > > > > > > > > Regards > > > > > Neil > > > > > > > > > > > > > > Its exported in the version map file: > per_lcore__lcore_id; > > Thanks Neil, I checked and both linux and bsd rte_eal_version.map have it. I compared .map file between "changed code" and the original, they are same for both linux and bsd. In fact you had ACK'd v4 version of this patch series and no major changes after that. Please let me know if I missed something. dpdk-common-eal-f-4 has all the common code movement changes dpdk-tmp is original code #/home/rkerur/dpdk-common-eal-f-4/dpdk# diff lib/librte_eal/linuxapp/eal/rte_eal_version.map /home/rkerur/dpdk-tmp/dpdk/lib/librte_eal/linuxapp/eal/rte_eal_version.map <No difference> #/home/rkerur/dpdk-common-eal-f-4/dpdk# diff lib/librte_eal/bsdapp/eal/rte_eal_version.map /home/rkerur/dpdk-tmp/dpdk/lib/librte_eal/bsdapp/eal/rte_eal_version.map <No difference> Thanks, Ravi Neil > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-24 19:21 ` Ravi Kerur @ 2015-04-24 19:51 ` Neil Horman 2015-04-24 21:24 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-24 19:51 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman <nhorman@tuxdriver.com> > > wrote: > > > > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman <nhorman@tuxdriver.com> > > > > wrote: > > > > > > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > > > > > > > Changes in v7 > > > > > > > Remove _setname_ pthread calls. > > > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > > > > > > > > > > > > Changes in v6 > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > > > > > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > > > > > > > respectively. Plan to use _getname_ in RTE_LOG when available. > > > > > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > > > > > > > > > > > > > Changes in v5 > > > > > > > Rebase to latest code. > > > > > > > > > > > > > > Changes in v4 > > > > > > > None > > > > > > > > > > > > > > Changes in v3 > > > > > > > Changed subject to be more explicit on file name inclusion. > > > > > > > > > > > > > > Changes in v2 > > > > > > > None > > > > > > > > > > > > > > Changes in v1 > > > > > > > eal_thread.c has minor differences between Linux and BSD, move > > > > > > > entire file into common directory. > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor differences. > > > > > > > Rename eal_thread.c to eal_common_thread.c > > > > > > > Makefile changes to reflect file move and name change. > > > > > > > Fix checkpatch warnings. > > > > > > > > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > > > > --- > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > > > > ------------------------------ > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > > > > ++++++++++++++++++++++++++++- > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > > > > +----------------------------- > > > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > > > > > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > index 2357cfa..55971b9 100644 > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > > > > > > > # workaround for a gcc bug with noreturn attribute > > > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > > > > endif > > > > > > > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > index 9a03437..5714b8f 100644 > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > @@ -35,163 +35,11 @@ > > > > > > > #include <stdio.h> > > > > > > > #include <stdlib.h> > > > > > > > #include <stdint.h> > > > > > > > -#include <unistd.h> > > > > > > > -#include <sched.h> > > > > > > > -#include <pthread_np.h> > > > > > > > -#include <sys/queue.h> > > > > > > > #include <sys/thr.h> > > > > > > > > > > > > > > -#include <rte_debug.h> > > > > > > > -#include <rte_atomic.h> > > > > > > > -#include <rte_launch.h> > > > > > > > -#include <rte_log.h> > > > > > > > -#include <rte_memory.h> > > > > > > > -#include <rte_memzone.h> > > > > > > > -#include <rte_per_lcore.h> > > > > > > > -#include <rte_eal.h> > > > > > > > -#include <rte_per_lcore.h> > > > > > > > -#include <rte_lcore.h> > > > > > > > - > > > > > > > #include "eal_private.h" > > > > > > > #include "eal_thread.h" > > > > > > > > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > > > > > > NAK, these are exported symbols, you can't remove them without > > going > > > > > > through the > > > > > > deprecation process. > > > > > > > > > > > > > > > > > They are not removed/deleted, they are moved from eal_thread.c to > > > > > eal_common_thread.c file since it is common to both Linux and BSD. > > > > > > > > > Then perhaps you forgot to export the symbol? Its showing up as > > removed > > > > on the > > > > ABI checker utility. > > > > > > > > Neil > > > > > > > > > > Can you please show me in the current code where it is being exported? I > > > have only moved definitions to _common_ files, not sure why it should be > > > exported now. I searched in the current code for RTE_DEFINE_PER_LCORE > > > > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * > > > app/test/test_per_lcore.c:static RTE_DEFINE_PER_LCORE(unsigned, test) = > > > 0x12345678; > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > _lcore_id) = LCORE_ID_ANY; > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > _cpuset); > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > _lcore_id) = LCORE_ID_ANY; > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > _cpuset); > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > > RTE_DEFINE_PER_LCORE(type, name) \ > > > lib/librte_eal/common/include/rte_eal.h: static > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > > _rte_errno); > > > lib/librte_eal/common/eal_common_errno.c: static > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > > > > > > > > Thanks > > > > > Ravi > > > > > > > > > > Regards > > > > > > Neil > > > > > > > > > > > > > > > > > > Its exported in the version map file: > > per_lcore__lcore_id; > > > > > Thanks Neil, I checked and both linux and bsd rte_eal_version.map have it. > I compared .map file between "changed code" and the original, they are same > for both linux and bsd. In fact you had ACK'd v4 version of this patch > series and no major changes after that. Please let me know if I missed > something. > I did, and I'm retracting that, because I didn't think to check the ABI compatibility on this. But I ran it throught the ABI checking script this and this error popped out. You should run it as well, its in the scripts directory. I see in your first patch you removed it and re-added it in the common section. But something about how its building is causing it to not show up as an exported symbol, which is problematic, as other applications are going to want access to it. It also possible that the ABI checker is throwing a false positive, but either way, it needs to be looked into prior to moving forward with this. Neil > dpdk-common-eal-f-4 has all the common code movement changes > dpdk-tmp is original code > > #/home/rkerur/dpdk-common-eal-f-4/dpdk# diff > lib/librte_eal/linuxapp/eal/rte_eal_version.map > /home/rkerur/dpdk-tmp/dpdk/lib/librte_eal/linuxapp/eal/rte_eal_version.map > > <No difference> > > #/home/rkerur/dpdk-common-eal-f-4/dpdk# diff > lib/librte_eal/bsdapp/eal/rte_eal_version.map > /home/rkerur/dpdk-tmp/dpdk/lib/librte_eal/bsdapp/eal/rte_eal_version.map > > <No difference> > > Thanks, > Ravi > > Neil > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-24 19:51 ` Neil Horman @ 2015-04-24 21:24 ` Ravi Kerur 2015-04-25 1:45 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-24 21:24 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman <nhorman@tuxdriver.com> wrote: > On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman <nhorman@tuxdriver.com> > wrote: > > > > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman <nhorman@tuxdriver.com> > > > wrote: > > > > > > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > nhorman@tuxdriver.com> > > > > > wrote: > > > > > > > > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > > > > > > > > Changes in v7 > > > > > > > > Remove _setname_ pthread calls. > > > > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > > > > > > > > > > > > > > Changes in v6 > > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > > > > > > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > > > > > > > > respectively. Plan to use _getname_ in RTE_LOG when > available. > > > > > > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > > > > > > > > > > > > > > > Changes in v5 > > > > > > > > Rebase to latest code. > > > > > > > > > > > > > > > > Changes in v4 > > > > > > > > None > > > > > > > > > > > > > > > > Changes in v3 > > > > > > > > Changed subject to be more explicit on file name inclusion. > > > > > > > > > > > > > > > > Changes in v2 > > > > > > > > None > > > > > > > > > > > > > > > > Changes in v1 > > > > > > > > eal_thread.c has minor differences between Linux and BSD, > move > > > > > > > > entire file into common directory. > > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor > differences. > > > > > > > > Rename eal_thread.c to eal_common_thread.c > > > > > > > > Makefile changes to reflect file move and name change. > > > > > > > > Fix checkpatch warnings. > > > > > > > > > > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > > > > > --- > > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > > > > > ------------------------------ > > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > > > > > ++++++++++++++++++++++++++++- > > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > > > > > +----------------------------- > > > > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > > > > > > > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > > index 2357cfa..55971b9 100644 > > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > > > > > > > > # workaround for a gcc bug with noreturn attribute > > > > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > > > > > endif > > > > > > > > > > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > > index 9a03437..5714b8f 100644 > > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > > @@ -35,163 +35,11 @@ > > > > > > > > #include <stdio.h> > > > > > > > > #include <stdlib.h> > > > > > > > > #include <stdint.h> > > > > > > > > -#include <unistd.h> > > > > > > > > -#include <sched.h> > > > > > > > > -#include <pthread_np.h> > > > > > > > > -#include <sys/queue.h> > > > > > > > > #include <sys/thr.h> > > > > > > > > > > > > > > > > -#include <rte_debug.h> > > > > > > > > -#include <rte_atomic.h> > > > > > > > > -#include <rte_launch.h> > > > > > > > > -#include <rte_log.h> > > > > > > > > -#include <rte_memory.h> > > > > > > > > -#include <rte_memzone.h> > > > > > > > > -#include <rte_per_lcore.h> > > > > > > > > -#include <rte_eal.h> > > > > > > > > -#include <rte_per_lcore.h> > > > > > > > > -#include <rte_lcore.h> > > > > > > > > - > > > > > > > > #include "eal_private.h" > > > > > > > > #include "eal_thread.h" > > > > > > > > > > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > > > > > > > NAK, these are exported symbols, you can't remove them without > > > going > > > > > > > through the > > > > > > > deprecation process. > > > > > > > > > > > > > > > > > > > > They are not removed/deleted, they are moved from eal_thread.c to > > > > > > eal_common_thread.c file since it is common to both Linux and > BSD. > > > > > > > > > > > Then perhaps you forgot to export the symbol? Its showing up as > > > removed > > > > > on the > > > > > ABI checker utility. > > > > > > > > > > Neil > > > > > > > > > > > > > Can you please show me in the current code where it is being > exported? I > > > > have only moved definitions to _common_ files, not sure why it > should be > > > > exported now. I searched in the current code for > RTE_DEFINE_PER_LCORE > > > > > > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * > > > > app/test/test_per_lcore.c:static RTE_DEFINE_PER_LCORE(unsigned, > test) = > > > > 0x12345678; > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > _lcore_id) = LCORE_ID_ANY; > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > _cpuset); > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > _lcore_id) = LCORE_ID_ANY; > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > _cpuset); > > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > > > RTE_DEFINE_PER_LCORE(type, name) \ > > > > lib/librte_eal/common/include/rte_eal.h: static > > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > > > _rte_errno); > > > > lib/librte_eal/common/eal_common_errno.c: static > > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > > > > > > > > > > > Thanks > > > > > > Ravi > > > > > > > > > > > > Regards > > > > > > > Neil > > > > > > > > > > > > > > > > > > > > > > Its exported in the version map file: > > > per_lcore__lcore_id; > > > > > > > > Thanks Neil, I checked and both linux and bsd rte_eal_version.map have > it. > > I compared .map file between "changed code" and the original, they are > same > > for both linux and bsd. In fact you had ACK'd v4 version of this patch > > series and no major changes after that. Please let me know if I missed > > something. > > > I did, and I'm retracting that, because I didn't think to check the ABI > compatibility on this. But I ran it throught the ABI checking script this > and > this error popped out. You should run it as well, its in the scripts > directory. > > > I see in your first patch you removed it and re-added it in the common > section. > But something about how its building is causing it to not show up as an > exported > symbol, which is problematic, as other applications are going to want > access to > it. > > It also possible that the ABI checker is throwing a false positive, but > either > way, it needs to be looked into prior to moving forward with this. > > I did following things. Put a tag (v2.0.0-before-common-eal) before EAL common functions changes for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: remove unneeded include) Put a tag (v2.0.0-common-eal) after EAL common functions changes for commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common functions in eal_pci.c) Ran validate-abi against x86_64-native-linuxapp-gcc and v2.0.0-rc3 and v2.0.0-before-common-eal, html report for librte_eal.so shows removed symbols for "per_lcore__cpuset" v2.0.0-rc3 and v2.0.0-common-eal, html report for librte_eal.so shows removed symbols for "per_lcore__cpuset" Removed symbol is different from what you have reported and in my case I see it even before my commit. If you are interested I can unicast you html report file. Please let me know how to proceed. > Neil > > > dpdk-common-eal-f-4 has all the common code movement changes > > dpdk-tmp is original code > > > > #/home/rkerur/dpdk-common-eal-f-4/dpdk# diff > > lib/librte_eal/linuxapp/eal/rte_eal_version.map > > > /home/rkerur/dpdk-tmp/dpdk/lib/librte_eal/linuxapp/eal/rte_eal_version.map > > > > <No difference> > > > > #/home/rkerur/dpdk-common-eal-f-4/dpdk# diff > > lib/librte_eal/bsdapp/eal/rte_eal_version.map > > /home/rkerur/dpdk-tmp/dpdk/lib/librte_eal/bsdapp/eal/rte_eal_version.map > > > > <No difference> > > > > Thanks, > > Ravi > > > > Neil > > > > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-24 21:24 ` Ravi Kerur @ 2015-04-25 1:45 ` Ravi Kerur 2015-04-25 12:32 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-25 1:45 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> wrote: > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman <nhorman@tuxdriver.com> > wrote: > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman <nhorman@tuxdriver.com> >> wrote: >> > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman <nhorman@tuxdriver.com >> > >> > > wrote: >> > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < >> nhorman@tuxdriver.com> >> > > > > wrote: >> > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: >> > > > > > > > Changes in v7 >> > > > > > > > Remove _setname_ pthread calls. >> > > > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. >> > > > > > > > >> > > > > > > > Changes in v6 >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG when >> available. >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. >> > > > > > > > >> > > > > > > > Changes in v5 >> > > > > > > > Rebase to latest code. >> > > > > > > > >> > > > > > > > Changes in v4 >> > > > > > > > None >> > > > > > > > >> > > > > > > > Changes in v3 >> > > > > > > > Changed subject to be more explicit on file name inclusion. >> > > > > > > > >> > > > > > > > Changes in v2 >> > > > > > > > None >> > > > > > > > >> > > > > > > > Changes in v1 >> > > > > > > > eal_thread.c has minor differences between Linux and BSD, >> move >> > > > > > > > entire file into common directory. >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor >> differences. >> > > > > > > > Rename eal_thread.c to eal_common_thread.c >> > > > > > > > Makefile changes to reflect file move and name change. >> > > > > > > > Fix checkpatch warnings. >> > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> >> > > > > > > > --- >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 >> > > > > > > ------------------------------ >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 >> > > > > > > ++++++++++++++++++++++++++++- >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 >> > > > > > > +----------------------------- >> > > > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) >> > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile >> > > > > > > > index 2357cfa..55971b9 100644 >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE >> > > > > > > > # workaround for a gcc bug with noreturn attribute >> > > > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type >> > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type >> > > > > > > > endif >> > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c >> > > > > > > > index 9a03437..5714b8f 100644 >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c >> > > > > > > > @@ -35,163 +35,11 @@ >> > > > > > > > #include <stdio.h> >> > > > > > > > #include <stdlib.h> >> > > > > > > > #include <stdint.h> >> > > > > > > > -#include <unistd.h> >> > > > > > > > -#include <sched.h> >> > > > > > > > -#include <pthread_np.h> >> > > > > > > > -#include <sys/queue.h> >> > > > > > > > #include <sys/thr.h> >> > > > > > > > >> > > > > > > > -#include <rte_debug.h> >> > > > > > > > -#include <rte_atomic.h> >> > > > > > > > -#include <rte_launch.h> >> > > > > > > > -#include <rte_log.h> >> > > > > > > > -#include <rte_memory.h> >> > > > > > > > -#include <rte_memzone.h> >> > > > > > > > -#include <rte_per_lcore.h> >> > > > > > > > -#include <rte_eal.h> >> > > > > > > > -#include <rte_per_lcore.h> >> > > > > > > > -#include <rte_lcore.h> >> > > > > > > > - >> > > > > > > > #include "eal_private.h" >> > > > > > > > #include "eal_thread.h" >> > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; >> > > > > > > NAK, these are exported symbols, you can't remove them without >> > > going >> > > > > > > through the >> > > > > > > deprecation process. >> > > > > > > >> > > > > > > >> > > > > > They are not removed/deleted, they are moved from eal_thread.c >> to >> > > > > > eal_common_thread.c file since it is common to both Linux and >> BSD. >> > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its showing up as >> > > removed >> > > > > on the >> > > > > ABI checker utility. >> > > > > >> > > > > Neil >> > > > > >> > > > >> > > > Can you please show me in the current code where it is being >> exported? I >> > > > have only moved definitions to _common_ files, not sure why it >> should be >> > > > exported now. I searched in the current code for >> RTE_DEFINE_PER_LCORE >> > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * >> > > > app/test/test_per_lcore.c:static RTE_DEFINE_PER_LCORE(unsigned, >> test) = >> > > > 0x12345678; >> > > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, >> > > > _lcore_id) = LCORE_ID_ANY; >> > > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; >> > > > >> > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, >> > > > _cpuset); >> > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, >> > > > _lcore_id) = LCORE_ID_ANY; >> > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; >> > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, >> > > > _cpuset); >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define >> > > > RTE_DEFINE_PER_LCORE(type, name) \ >> > > > lib/librte_eal/common/include/rte_eal.h: static >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; >> > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, >> > > > _rte_errno); >> > > > lib/librte_eal/common/eal_common_errno.c: static >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); >> > > > >> > > > >> > > > > > Thanks >> > > > > > Ravi >> > > > > > >> > > > > > Regards >> > > > > > > Neil >> > > > > > > >> > > > > > > >> > > > > >> > > Its exported in the version map file: >> > > per_lcore__lcore_id; >> > > >> > > >> > Thanks Neil, I checked and both linux and bsd rte_eal_version.map have >> it. >> > I compared .map file between "changed code" and the original, they are >> same >> > for both linux and bsd. In fact you had ACK'd v4 version of this patch >> > series and no major changes after that. Please let me know if I missed >> > something. >> > >> I did, and I'm retracting that, because I didn't think to check the ABI >> compatibility on this. But I ran it throught the ABI checking script >> this and >> this error popped out. You should run it as well, its in the scripts >> directory. >> >> >> I see in your first patch you removed it and re-added it in the common >> section. >> But something about how its building is causing it to not show up as an >> exported >> symbol, which is problematic, as other applications are going to want >> access to >> it. >> >> It also possible that the ABI checker is throwing a false positive, but >> either >> way, it needs to be looked into prior to moving forward with this. >> >> > I did following things. > > Put a tag (v2.0.0-before-common-eal) before EAL common functions changes > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: remove unneeded > include) > > Put a tag (v2.0.0-common-eal) after EAL common functions changes for > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common functions in > eal_pci.c) > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for librte_eal.so > shows removed symbols for "per_lcore__cpuset" > > v2.0.0-rc3 and v2.0.0-common-eal, html report for librte_eal.so shows > removed symbols for "per_lcore__cpuset" > > Removed symbol is different from what you have reported and in my case I > see it even before my commit. If you are interested I can unicast you html > report file. Please let me know how to proceed. > > I did some experiment and found some interesting things. I will take eal.c as an example eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c eal_common_proc_type.c and eal_common_app_usage.c. In linuxapp/eal/Makefile if I compile new files right after eal.c as shown below SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c ... validate-abi results matches baseline. Instead if i place new _common_ files in common area in linuxapp/eal/Makefile as shown below # from common dir SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c ... validate-abi reports problem in binary compatibility and source compatiblity eal_filesystem.h, librte_eal.so.1 [+] eal_parse_sysfs_value ( char const* filename, unsigned long* val ) @@ DPDK_2.0 (2) I believe files in common and linuxapp directory are compiled same way so not sure why placement in makefile makes difference. Could this be false-positive from validate-abi script?? > Neil >> >> > dpdk-common-eal-f-4 has all the common code movement changes >> > dpdk-tmp is original code >> > >> > #/home/rkerur/dpdk-common-eal-f-4/dpdk# diff >> > lib/librte_eal/linuxapp/eal/rte_eal_version.map >> > >> /home/rkerur/dpdk-tmp/dpdk/lib/librte_eal/linuxapp/eal/rte_eal_version.map >> > >> > <No difference> >> > >> > #/home/rkerur/dpdk-common-eal-f-4/dpdk# diff >> > lib/librte_eal/bsdapp/eal/rte_eal_version.map >> > /home/rkerur/dpdk-tmp/dpdk/lib/librte_eal/bsdapp/eal/rte_eal_version.map >> > >> > <No difference> >> > >> > Thanks, >> > Ravi >> > >> > Neil >> > > >> > > >> > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-25 1:45 ` Ravi Kerur @ 2015-04-25 12:32 ` Neil Horman 2015-04-25 13:02 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-25 12:32 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Fri, Apr 24, 2015 at 06:45:06PM -0700, Ravi Kerur wrote: > On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> wrote: > > > > > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman <nhorman@tuxdriver.com> > > wrote: > > > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman <nhorman@tuxdriver.com> > >> wrote: > >> > > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman <nhorman@tuxdriver.com > >> > > >> > > wrote: > >> > > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > >> nhorman@tuxdriver.com> > >> > > > > wrote: > >> > > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > >> > > > > > > > Changes in v7 > >> > > > > > > > Remove _setname_ pthread calls. > >> > > > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > >> > > > > > > > > >> > > > > > > > Changes in v6 > >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG when > >> available. > >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > >> > > > > > > > > >> > > > > > > > Changes in v5 > >> > > > > > > > Rebase to latest code. > >> > > > > > > > > >> > > > > > > > Changes in v4 > >> > > > > > > > None > >> > > > > > > > > >> > > > > > > > Changes in v3 > >> > > > > > > > Changed subject to be more explicit on file name inclusion. > >> > > > > > > > > >> > > > > > > > Changes in v2 > >> > > > > > > > None > >> > > > > > > > > >> > > > > > > > Changes in v1 > >> > > > > > > > eal_thread.c has minor differences between Linux and BSD, > >> move > >> > > > > > > > entire file into common directory. > >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor > >> differences. > >> > > > > > > > Rename eal_thread.c to eal_common_thread.c > >> > > > > > > > Makefile changes to reflect file move and name change. > >> > > > > > > > Fix checkpatch warnings. > >> > > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > >> > > > > > > > --- > >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > >> > > > > > > ------------------------------ > >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > >> > > > > > > ++++++++++++++++++++++++++++- > >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > >> > > > > > > +----------------------------- > >> > > > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > >> > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > >> > > > > > > > index 2357cfa..55971b9 100644 > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > >> > > > > > > > # workaround for a gcc bug with noreturn attribute > >> > > > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > >> > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > >> > > > > > > > endif > >> > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > >> > > > > > > > index 9a03437..5714b8f 100644 > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > >> > > > > > > > @@ -35,163 +35,11 @@ > >> > > > > > > > #include <stdio.h> > >> > > > > > > > #include <stdlib.h> > >> > > > > > > > #include <stdint.h> > >> > > > > > > > -#include <unistd.h> > >> > > > > > > > -#include <sched.h> > >> > > > > > > > -#include <pthread_np.h> > >> > > > > > > > -#include <sys/queue.h> > >> > > > > > > > #include <sys/thr.h> > >> > > > > > > > > >> > > > > > > > -#include <rte_debug.h> > >> > > > > > > > -#include <rte_atomic.h> > >> > > > > > > > -#include <rte_launch.h> > >> > > > > > > > -#include <rte_log.h> > >> > > > > > > > -#include <rte_memory.h> > >> > > > > > > > -#include <rte_memzone.h> > >> > > > > > > > -#include <rte_per_lcore.h> > >> > > > > > > > -#include <rte_eal.h> > >> > > > > > > > -#include <rte_per_lcore.h> > >> > > > > > > > -#include <rte_lcore.h> > >> > > > > > > > - > >> > > > > > > > #include "eal_private.h" > >> > > > > > > > #include "eal_thread.h" > >> > > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > >> > > > > > > NAK, these are exported symbols, you can't remove them without > >> > > going > >> > > > > > > through the > >> > > > > > > deprecation process. > >> > > > > > > > >> > > > > > > > >> > > > > > They are not removed/deleted, they are moved from eal_thread.c > >> to > >> > > > > > eal_common_thread.c file since it is common to both Linux and > >> BSD. > >> > > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its showing up as > >> > > removed > >> > > > > on the > >> > > > > ABI checker utility. > >> > > > > > >> > > > > Neil > >> > > > > > >> > > > > >> > > > Can you please show me in the current code where it is being > >> exported? I > >> > > > have only moved definitions to _common_ files, not sure why it > >> should be > >> > > > exported now. I searched in the current code for > >> RTE_DEFINE_PER_LCORE > >> > > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * > >> > > > app/test/test_per_lcore.c:static RTE_DEFINE_PER_LCORE(unsigned, > >> test) = > >> > > > 0x12345678; > >> > > > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > >> > > > _lcore_id) = LCORE_ID_ANY; > >> > > > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > >> > > > > >> > > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > >> > > > _cpuset); > >> > > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > >> > > > _lcore_id) = LCORE_ID_ANY; > >> > > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > >> > > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > >> > > > _cpuset); > >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > >> > > > RTE_DEFINE_PER_LCORE(type, name) \ > >> > > > lib/librte_eal/common/include/rte_eal.h: static > >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > >> > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > >> > > > _rte_errno); > >> > > > lib/librte_eal/common/eal_common_errno.c: static > >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > >> > > > > >> > > > > >> > > > > > Thanks > >> > > > > > Ravi > >> > > > > > > >> > > > > > Regards > >> > > > > > > Neil > >> > > > > > > > >> > > > > > > > >> > > > > > >> > > Its exported in the version map file: > >> > > per_lcore__lcore_id; > >> > > > >> > > > >> > Thanks Neil, I checked and both linux and bsd rte_eal_version.map have > >> it. > >> > I compared .map file between "changed code" and the original, they are > >> same > >> > for both linux and bsd. In fact you had ACK'd v4 version of this patch > >> > series and no major changes after that. Please let me know if I missed > >> > something. > >> > > >> I did, and I'm retracting that, because I didn't think to check the ABI > >> compatibility on this. But I ran it throught the ABI checking script > >> this and > >> this error popped out. You should run it as well, its in the scripts > >> directory. > >> > >> > >> I see in your first patch you removed it and re-added it in the common > >> section. > >> But something about how its building is causing it to not show up as an > >> exported > >> symbol, which is problematic, as other applications are going to want > >> access to > >> it. > >> > >> It also possible that the ABI checker is throwing a false positive, but > >> either > >> way, it needs to be looked into prior to moving forward with this. > >> > >> > > I did following things. > > > > Put a tag (v2.0.0-before-common-eal) before EAL common functions changes > > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: remove unneeded > > include) > > > > Put a tag (v2.0.0-common-eal) after EAL common functions changes for > > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common functions in > > eal_pci.c) > > > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for librte_eal.so > > shows removed symbols for "per_lcore__cpuset" > > > > v2.0.0-rc3 and v2.0.0-common-eal, html report for librte_eal.so shows > > removed symbols for "per_lcore__cpuset" > > > > Removed symbol is different from what you have reported and in my case I > > see it even before my commit. If you are interested I can unicast you html > > report file. Please let me know how to proceed. > > > > > > I did some experiment and found some interesting things. I will take eal.c > as an example > > eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c > eal_common_proc_type.c and eal_common_app_usage.c. In linuxapp/eal/Makefile > if I compile new files right after eal.c as shown below > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c > ... > > validate-abi results matches baseline. Instead if i place new _common_ > files in common area in linuxapp/eal/Makefile as shown below > > # from common dir > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c > ... > > validate-abi reports problem in binary compatibility and source > compatiblity > > eal_filesystem.h, librte_eal.so.1 > [+] eal_parse_sysfs_value ( char const* filename, unsigned long* val ) > @@ DPDK_2.0 (2) > > I believe files in common and linuxapp directory are compiled same way so > not sure why placement in makefile makes difference. > > Could this be false-positive from validate-abi script?? > It could be, yes. Though I'm more inclined to think that perhaps in the new version of the code we're not generating ithe same dwarf information out of it. In fact for some reason, I've checked both the build before and after your patch series, and the exported CFLAGS aren't getting passed to the build properly, implying that we're not building all the code in the validator with the -g flag, which the validator need to function properly. I'm looking into that Neil ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-25 12:32 ` Neil Horman @ 2015-04-25 13:02 ` Neil Horman 2015-04-26 0:09 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-25 13:02 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Sat, Apr 25, 2015 at 08:32:42AM -0400, Neil Horman wrote: > On Fri, Apr 24, 2015 at 06:45:06PM -0700, Ravi Kerur wrote: > > On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> wrote: > > > > > > > > > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman <nhorman@tuxdriver.com> > > > wrote: > > > > > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > > >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman <nhorman@tuxdriver.com> > > >> wrote: > > >> > > > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > > >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman <nhorman@tuxdriver.com > > >> > > > >> > > wrote: > > >> > > > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > > >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > > >> nhorman@tuxdriver.com> > > >> > > > > wrote: > > >> > > > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur wrote: > > >> > > > > > > > Changes in v7 > > >> > > > > > > > Remove _setname_ pthread calls. > > >> > > > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > >> > > > > > > > > > >> > > > > > > > Changes in v6 > > >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c file. > > >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for Linux/FreeBSD > > >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG when > > >> available. > > >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to print thread_id. > > >> > > > > > > > > > >> > > > > > > > Changes in v5 > > >> > > > > > > > Rebase to latest code. > > >> > > > > > > > > > >> > > > > > > > Changes in v4 > > >> > > > > > > > None > > >> > > > > > > > > > >> > > > > > > > Changes in v3 > > >> > > > > > > > Changed subject to be more explicit on file name inclusion. > > >> > > > > > > > > > >> > > > > > > > Changes in v2 > > >> > > > > > > > None > > >> > > > > > > > > > >> > > > > > > > Changes in v1 > > >> > > > > > > > eal_thread.c has minor differences between Linux and BSD, > > >> move > > >> > > > > > > > entire file into common directory. > > >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor > > >> differences. > > >> > > > > > > > Rename eal_thread.c to eal_common_thread.c > > >> > > > > > > > Makefile changes to reflect file move and name change. > > >> > > > > > > > Fix checkpatch warnings. > > >> > > > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > >> > > > > > > > --- > > >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > >> > > > > > > ------------------------------ > > >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > >> > > > > > > ++++++++++++++++++++++++++++- > > >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > >> > > > > > > +----------------------------- > > >> > > > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > >> > > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > >> > > > > > > > index 2357cfa..55971b9 100644 > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := -D_GNU_SOURCE > > >> > > > > > > > # workaround for a gcc bug with noreturn attribute > > >> > > > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > >> > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > >> > > > > > > > endif > > >> > > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > >> > > > > > > > index 9a03437..5714b8f 100644 > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > >> > > > > > > > @@ -35,163 +35,11 @@ > > >> > > > > > > > #include <stdio.h> > > >> > > > > > > > #include <stdlib.h> > > >> > > > > > > > #include <stdint.h> > > >> > > > > > > > -#include <unistd.h> > > >> > > > > > > > -#include <sched.h> > > >> > > > > > > > -#include <pthread_np.h> > > >> > > > > > > > -#include <sys/queue.h> > > >> > > > > > > > #include <sys/thr.h> > > >> > > > > > > > > > >> > > > > > > > -#include <rte_debug.h> > > >> > > > > > > > -#include <rte_atomic.h> > > >> > > > > > > > -#include <rte_launch.h> > > >> > > > > > > > -#include <rte_log.h> > > >> > > > > > > > -#include <rte_memory.h> > > >> > > > > > > > -#include <rte_memzone.h> > > >> > > > > > > > -#include <rte_per_lcore.h> > > >> > > > > > > > -#include <rte_eal.h> > > >> > > > > > > > -#include <rte_per_lcore.h> > > >> > > > > > > > -#include <rte_lcore.h> > > >> > > > > > > > - > > >> > > > > > > > #include "eal_private.h" > > >> > > > > > > > #include "eal_thread.h" > > >> > > > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = LCORE_ID_ANY; > > >> > > > > > > NAK, these are exported symbols, you can't remove them without > > >> > > going > > >> > > > > > > through the > > >> > > > > > > deprecation process. > > >> > > > > > > > > >> > > > > > > > > >> > > > > > They are not removed/deleted, they are moved from eal_thread.c > > >> to > > >> > > > > > eal_common_thread.c file since it is common to both Linux and > > >> BSD. > > >> > > > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its showing up as > > >> > > removed > > >> > > > > on the > > >> > > > > ABI checker utility. > > >> > > > > > > >> > > > > Neil > > >> > > > > > > >> > > > > > >> > > > Can you please show me in the current code where it is being > > >> exported? I > > >> > > > have only moved definitions to _common_ files, not sure why it > > >> should be > > >> > > > exported now. I searched in the current code for > > >> RTE_DEFINE_PER_LCORE > > >> > > > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * > > >> > > > app/test/test_per_lcore.c:static RTE_DEFINE_PER_LCORE(unsigned, > > >> test) = > > >> > > > 0x12345678; > > >> > > > > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > >> > > > _lcore_id) = LCORE_ID_ANY; > > >> > > > > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > >> > > > > > >> > > > > >> lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > >> > > > _cpuset); > > >> > > > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > >> > > > _lcore_id) = LCORE_ID_ANY; > > >> > > > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > >> > > > > > >> lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > >> > > > _cpuset); > > >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > >> > > > RTE_DEFINE_PER_LCORE(type, name) \ > > >> > > > lib/librte_eal/common/include/rte_eal.h: static > > >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > >> > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > >> > > > _rte_errno); > > >> > > > lib/librte_eal/common/eal_common_errno.c: static > > >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > >> > > > > > >> > > > > > >> > > > > > Thanks > > >> > > > > > Ravi > > >> > > > > > > > >> > > > > > Regards > > >> > > > > > > Neil > > >> > > > > > > > > >> > > > > > > > > >> > > > > > > >> > > Its exported in the version map file: > > >> > > per_lcore__lcore_id; > > >> > > > > >> > > > > >> > Thanks Neil, I checked and both linux and bsd rte_eal_version.map have > > >> it. > > >> > I compared .map file between "changed code" and the original, they are > > >> same > > >> > for both linux and bsd. In fact you had ACK'd v4 version of this patch > > >> > series and no major changes after that. Please let me know if I missed > > >> > something. > > >> > > > >> I did, and I'm retracting that, because I didn't think to check the ABI > > >> compatibility on this. But I ran it throught the ABI checking script > > >> this and > > >> this error popped out. You should run it as well, its in the scripts > > >> directory. > > >> > > >> > > >> I see in your first patch you removed it and re-added it in the common > > >> section. > > >> But something about how its building is causing it to not show up as an > > >> exported > > >> symbol, which is problematic, as other applications are going to want > > >> access to > > >> it. > > >> > > >> It also possible that the ABI checker is throwing a false positive, but > > >> either > > >> way, it needs to be looked into prior to moving forward with this. > > >> > > >> > > > I did following things. > > > > > > Put a tag (v2.0.0-before-common-eal) before EAL common functions changes > > > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: remove unneeded > > > include) > > > > > > Put a tag (v2.0.0-common-eal) after EAL common functions changes for > > > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common functions in > > > eal_pci.c) > > > > > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > > > > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for librte_eal.so > > > shows removed symbols for "per_lcore__cpuset" > > > > > > v2.0.0-rc3 and v2.0.0-common-eal, html report for librte_eal.so shows > > > removed symbols for "per_lcore__cpuset" > > > > > > Removed symbol is different from what you have reported and in my case I > > > see it even before my commit. If you are interested I can unicast you html > > > report file. Please let me know how to proceed. > > > > > > > > > > I did some experiment and found some interesting things. I will take eal.c > > as an example > > > > eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c > > eal_common_proc_type.c and eal_common_app_usage.c. In linuxapp/eal/Makefile > > if I compile new files right after eal.c as shown below > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c > > ... > > > > validate-abi results matches baseline. Instead if i place new _common_ > > files in common area in linuxapp/eal/Makefile as shown below > > > > # from common dir > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c > > ... > > > > validate-abi reports problem in binary compatibility and source > > compatiblity > > > > eal_filesystem.h, librte_eal.so.1 > > [+] eal_parse_sysfs_value ( char const* filename, unsigned long* val ) > > @@ DPDK_2.0 (2) > > > > I believe files in common and linuxapp directory are compiled same way so > > not sure why placement in makefile makes difference. > > > > Could this be false-positive from validate-abi script?? > > > It could be, yes. Though I'm more inclined to think that perhaps in the new > version of the code we're not generating ithe same dwarf information out of it. > In fact for some reason, I've checked both the build before and after your > patch series, and the exported CFLAGS aren't getting passed to the build > properly, implying that we're not building all the code in the validator with > the -g flag, which the validator need to function properly. I'm looking into > that > Neil > > Found the problem, I was stupidly reading the report incorrectly. The problem regarding _lcore_id is a source compatibilty issue (because the symbol moved to a new location), which is irrelevant to us. Its not in any way a binary compat problem, which is what we care about. Sorry for the noise. I do still have a few concerns about some changed calling conventions with a few other functions, which I'll look into on monday. Regards Neil ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-25 13:02 ` Neil Horman @ 2015-04-26 0:09 ` Ravi Kerur 2015-04-27 13:44 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-26 0:09 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Sat, Apr 25, 2015 at 6:02 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > On Sat, Apr 25, 2015 at 08:32:42AM -0400, Neil Horman wrote: > > On Fri, Apr 24, 2015 at 06:45:06PM -0700, Ravi Kerur wrote: > > > On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> wrote: > > > > > > > > > > > > > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman <nhorman@tuxdriver.com > > > > > > wrote: > > > > > > > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > > > >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman < > nhorman@tuxdriver.com> > > > >> wrote: > > > >> > > > > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > > > >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman < > nhorman@tuxdriver.com > > > >> > > > > >> > > wrote: > > > >> > > > > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > > > >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > > > >> nhorman@tuxdriver.com> > > > >> > > > > wrote: > > > >> > > > > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur > wrote: > > > >> > > > > > > > Changes in v7 > > > >> > > > > > > > Remove _setname_ pthread calls. > > > >> > > > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > >> > > > > > > > > > > >> > > > > > > > Changes in v6 > > > >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c > file. > > > >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for > Linux/FreeBSD > > > >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG when > > > >> available. > > > >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to print > thread_id. > > > >> > > > > > > > > > > >> > > > > > > > Changes in v5 > > > >> > > > > > > > Rebase to latest code. > > > >> > > > > > > > > > > >> > > > > > > > Changes in v4 > > > >> > > > > > > > None > > > >> > > > > > > > > > > >> > > > > > > > Changes in v3 > > > >> > > > > > > > Changed subject to be more explicit on file name > inclusion. > > > >> > > > > > > > > > > >> > > > > > > > Changes in v2 > > > >> > > > > > > > None > > > >> > > > > > > > > > > >> > > > > > > > Changes in v1 > > > >> > > > > > > > eal_thread.c has minor differences between Linux and > BSD, > > > >> move > > > >> > > > > > > > entire file into common directory. > > > >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor > > > >> differences. > > > >> > > > > > > > Rename eal_thread.c to eal_common_thread.c > > > >> > > > > > > > Makefile changes to reflect file move and name change. > > > >> > > > > > > > Fix checkpatch warnings. > > > >> > > > > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > >> > > > > > > > --- > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > >> > > > > > > ------------------------------ > > > >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > >> > > > > > > ++++++++++++++++++++++++++++- > > > >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > >> > > > > > > +----------------------------- > > > >> > > > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > >> > > > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > >> > > > > > > > index 2357cfa..55971b9 100644 > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := > -D_GNU_SOURCE > > > >> > > > > > > > # workaround for a gcc bug with noreturn attribute > > > >> > > > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > >> > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > >> > > > > > > > endif > > > >> > > > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > >> > > > > > > > index 9a03437..5714b8f 100644 > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > >> > > > > > > > @@ -35,163 +35,11 @@ > > > >> > > > > > > > #include <stdio.h> > > > >> > > > > > > > #include <stdlib.h> > > > >> > > > > > > > #include <stdint.h> > > > >> > > > > > > > -#include <unistd.h> > > > >> > > > > > > > -#include <sched.h> > > > >> > > > > > > > -#include <pthread_np.h> > > > >> > > > > > > > -#include <sys/queue.h> > > > >> > > > > > > > #include <sys/thr.h> > > > >> > > > > > > > > > > >> > > > > > > > -#include <rte_debug.h> > > > >> > > > > > > > -#include <rte_atomic.h> > > > >> > > > > > > > -#include <rte_launch.h> > > > >> > > > > > > > -#include <rte_log.h> > > > >> > > > > > > > -#include <rte_memory.h> > > > >> > > > > > > > -#include <rte_memzone.h> > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > >> > > > > > > > -#include <rte_eal.h> > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > >> > > > > > > > -#include <rte_lcore.h> > > > >> > > > > > > > - > > > >> > > > > > > > #include "eal_private.h" > > > >> > > > > > > > #include "eal_thread.h" > > > >> > > > > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = > LCORE_ID_ANY; > > > >> > > > > > > NAK, these are exported symbols, you can't remove them > without > > > >> > > going > > > >> > > > > > > through the > > > >> > > > > > > deprecation process. > > > >> > > > > > > > > > >> > > > > > > > > > >> > > > > > They are not removed/deleted, they are moved from > eal_thread.c > > > >> to > > > >> > > > > > eal_common_thread.c file since it is common to both Linux > and > > > >> BSD. > > > >> > > > > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its showing > up as > > > >> > > removed > > > >> > > > > on the > > > >> > > > > ABI checker utility. > > > >> > > > > > > > >> > > > > Neil > > > >> > > > > > > > >> > > > > > > >> > > > Can you please show me in the current code where it is being > > > >> exported? I > > > >> > > > have only moved definitions to _common_ files, not sure why it > > > >> should be > > > >> > > > exported now. I searched in the current code for > > > >> RTE_DEFINE_PER_LCORE > > > >> > > > > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * > > > >> > > > app/test/test_per_lcore.c:static > RTE_DEFINE_PER_LCORE(unsigned, > > > >> test) = > > > >> > > > 0x12345678; > > > >> > > > > > > >> > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > >> > > > > > > >> > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > >> > > > > > > >> > > > > > >> > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > >> > > > _cpuset); > > > >> > > > > > > >> > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > >> > > > > > > >> > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > >> > > > > > > >> > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > >> > > > _cpuset); > > > >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > > >> > > > RTE_DEFINE_PER_LCORE(type, name) \ > > > >> > > > lib/librte_eal/common/include/rte_eal.h: static > > > >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > > >> > > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > > >> > > > _rte_errno); > > > >> > > > lib/librte_eal/common/eal_common_errno.c: static > > > >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > >> > > > > > > >> > > > > > > >> > > > > > Thanks > > > >> > > > > > Ravi > > > >> > > > > > > > > >> > > > > > Regards > > > >> > > > > > > Neil > > > >> > > > > > > > > > >> > > > > > > > > > >> > > > > > > > >> > > Its exported in the version map file: > > > >> > > per_lcore__lcore_id; > > > >> > > > > > >> > > > > > >> > Thanks Neil, I checked and both linux and bsd rte_eal_version.map > have > > > >> it. > > > >> > I compared .map file between "changed code" and the original, > they are > > > >> same > > > >> > for both linux and bsd. In fact you had ACK'd v4 version of this > patch > > > >> > series and no major changes after that. Please let me know if I > missed > > > >> > something. > > > >> > > > > >> I did, and I'm retracting that, because I didn't think to check the > ABI > > > >> compatibility on this. But I ran it throught the ABI checking > script > > > >> this and > > > >> this error popped out. You should run it as well, its in the > scripts > > > >> directory. > > > >> > > > >> > > > >> I see in your first patch you removed it and re-added it in the > common > > > >> section. > > > >> But something about how its building is causing it to not show up > as an > > > >> exported > > > >> symbol, which is problematic, as other applications are going to > want > > > >> access to > > > >> it. > > > >> > > > >> It also possible that the ABI checker is throwing a false positive, > but > > > >> either > > > >> way, it needs to be looked into prior to moving forward with this. > > > >> > > > >> > > > > I did following things. > > > > > > > > Put a tag (v2.0.0-before-common-eal) before EAL common functions > changes > > > > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: remove > unneeded > > > > include) > > > > > > > > Put a tag (v2.0.0-common-eal) after EAL common functions changes for > > > > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common > functions in > > > > eal_pci.c) > > > > > > > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > > > > > > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for > librte_eal.so > > > > shows removed symbols for "per_lcore__cpuset" > > > > > > > > v2.0.0-rc3 and v2.0.0-common-eal, html report for librte_eal.so shows > > > > removed symbols for "per_lcore__cpuset" > > > > > > > > Removed symbol is different from what you have reported and in my > case I > > > > see it even before my commit. If you are interested I can unicast > you html > > > > report file. Please let me know how to proceed. > > > > > > > > > > > > > > I did some experiment and found some interesting things. I will take > eal.c > > > as an example > > > > > > eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c > > > eal_common_proc_type.c and eal_common_app_usage.c. In > linuxapp/eal/Makefile > > > if I compile new files right after eal.c as shown below > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c > > > ... > > > > > > validate-abi results matches baseline. Instead if i place new _common_ > > > files in common area in linuxapp/eal/Makefile as shown below > > > > > > # from common dir > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c > > > ... > > > > > > validate-abi reports problem in binary compatibility and source > > > compatiblity > > > > > > eal_filesystem.h, librte_eal.so.1 > > > [+] eal_parse_sysfs_value ( char const* filename, unsigned long* val ) > > > @@ DPDK_2.0 (2) > > > > > > I believe files in common and linuxapp directory are compiled same way > so > > > not sure why placement in makefile makes difference. > > > > > > Could this be false-positive from validate-abi script?? > > > > > It could be, yes. Though I'm more inclined to think that perhaps in the > new > > version of the code we're not generating ithe same dwarf information out > of it. > > In fact for some reason, I've checked both the build before and after > your > > patch series, and the exported CFLAGS aren't getting passed to the build > > properly, implying that we're not building all the code in the validator > with > > the -g flag, which the validator need to function properly. I'm looking > into > > that > > Neil > > > > > Found the problem, I was stupidly reading the report incorrectly. The > problem > regarding _lcore_id is a source compatibilty issue (because the symbol > moved to > a new location), which is irrelevant to us. Its not in any way a binary > compat > problem, which is what we care about. Sorry for the noise. > > I do still have a few concerns about some changed calling conventions with > a few > other functions, which I'll look into on monday. > > Please let me know your inputs on changed calling conventions. Most of them can be fixed by re-arranging moved code in _common_ files and order of compilation. Thanks, Ravi Regards > Neil > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-26 0:09 ` Ravi Kerur @ 2015-04-27 13:44 ` Neil Horman 2015-04-27 22:39 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-27 13:44 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Sat, Apr 25, 2015 at 05:09:01PM -0700, Ravi Kerur wrote: > On Sat, Apr 25, 2015 at 6:02 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > > > On Sat, Apr 25, 2015 at 08:32:42AM -0400, Neil Horman wrote: > > > On Fri, Apr 24, 2015 at 06:45:06PM -0700, Ravi Kerur wrote: > > > > On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> wrote: > > > > > > > > > > > > > > > > > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman <nhorman@tuxdriver.com > > > > > > > > wrote: > > > > > > > > > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > > > > >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman < > > nhorman@tuxdriver.com> > > > > >> wrote: > > > > >> > > > > > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > > > > >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman < > > nhorman@tuxdriver.com > > > > >> > > > > > >> > > wrote: > > > > >> > > > > > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur wrote: > > > > >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > > > > >> nhorman@tuxdriver.com> > > > > >> > > > > wrote: > > > > >> > > > > > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur > > wrote: > > > > >> > > > > > > > Changes in v7 > > > > >> > > > > > > > Remove _setname_ pthread calls. > > > > >> > > > > > > > Use rte_gettid() API in RTE_LOG to print thread_id. > > > > >> > > > > > > > > > > > >> > > > > > > > Changes in v6 > > > > >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from eal_common_thread.c > > file. > > > > >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for > > Linux/FreeBSD > > > > >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG when > > > > >> available. > > > > >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to print > > thread_id. > > > > >> > > > > > > > > > > > >> > > > > > > > Changes in v5 > > > > >> > > > > > > > Rebase to latest code. > > > > >> > > > > > > > > > > > >> > > > > > > > Changes in v4 > > > > >> > > > > > > > None > > > > >> > > > > > > > > > > > >> > > > > > > > Changes in v3 > > > > >> > > > > > > > Changed subject to be more explicit on file name > > inclusion. > > > > >> > > > > > > > > > > > >> > > > > > > > Changes in v2 > > > > >> > > > > > > > None > > > > >> > > > > > > > > > > > >> > > > > > > > Changes in v1 > > > > >> > > > > > > > eal_thread.c has minor differences between Linux and > > BSD, > > > > >> move > > > > >> > > > > > > > entire file into common directory. > > > > >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor > > > > >> differences. > > > > >> > > > > > > > Rename eal_thread.c to eal_common_thread.c > > > > >> > > > > > > > Makefile changes to reflect file move and name change. > > > > >> > > > > > > > Fix checkpatch warnings. > > > > >> > > > > > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > >> > > > > > > > --- > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 +- > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > > >> > > > > > > ------------------------------ > > > > >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > > >> > > > > > > ++++++++++++++++++++++++++++- > > > > >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > > >> > > > > > > +----------------------------- > > > > >> > > > > > > > 4 files changed, 148 insertions(+), 305 deletions(-) > > > > >> > > > > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > >> > > > > > > > index 2357cfa..55971b9 100644 > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := > > -D_GNU_SOURCE > > > > >> > > > > > > > # workaround for a gcc bug with noreturn attribute > > > > >> > > > > > > > # http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > >> > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > > >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > >> > > > > > > > endif > > > > >> > > > > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > >> > > > > > > > index 9a03437..5714b8f 100644 > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > >> > > > > > > > @@ -35,163 +35,11 @@ > > > > >> > > > > > > > #include <stdio.h> > > > > >> > > > > > > > #include <stdlib.h> > > > > >> > > > > > > > #include <stdint.h> > > > > >> > > > > > > > -#include <unistd.h> > > > > >> > > > > > > > -#include <sched.h> > > > > >> > > > > > > > -#include <pthread_np.h> > > > > >> > > > > > > > -#include <sys/queue.h> > > > > >> > > > > > > > #include <sys/thr.h> > > > > >> > > > > > > > > > > > >> > > > > > > > -#include <rte_debug.h> > > > > >> > > > > > > > -#include <rte_atomic.h> > > > > >> > > > > > > > -#include <rte_launch.h> > > > > >> > > > > > > > -#include <rte_log.h> > > > > >> > > > > > > > -#include <rte_memory.h> > > > > >> > > > > > > > -#include <rte_memzone.h> > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > >> > > > > > > > -#include <rte_eal.h> > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > >> > > > > > > > -#include <rte_lcore.h> > > > > >> > > > > > > > - > > > > >> > > > > > > > #include "eal_private.h" > > > > >> > > > > > > > #include "eal_thread.h" > > > > >> > > > > > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = > > LCORE_ID_ANY; > > > > >> > > > > > > NAK, these are exported symbols, you can't remove them > > without > > > > >> > > going > > > > >> > > > > > > through the > > > > >> > > > > > > deprecation process. > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > > > > > They are not removed/deleted, they are moved from > > eal_thread.c > > > > >> to > > > > >> > > > > > eal_common_thread.c file since it is common to both Linux > > and > > > > >> BSD. > > > > >> > > > > > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its showing > > up as > > > > >> > > removed > > > > >> > > > > on the > > > > >> > > > > ABI checker utility. > > > > >> > > > > > > > > >> > > > > Neil > > > > >> > > > > > > > > >> > > > > > > > >> > > > Can you please show me in the current code where it is being > > > > >> exported? I > > > > >> > > > have only moved definitions to _common_ files, not sure why it > > > > >> should be > > > > >> > > > exported now. I searched in the current code for > > > > >> RTE_DEFINE_PER_LCORE > > > > >> > > > > > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE * > > > > >> > > > app/test/test_per_lcore.c:static > > RTE_DEFINE_PER_LCORE(unsigned, > > > > >> test) = > > > > >> > > > 0x12345678; > > > > >> > > > > > > > >> > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > >> > > > > > > > >> > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > >> > > > > > > > >> > > > > > > >> > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > >> > > > _cpuset); > > > > >> > > > > > > > >> > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > >> > > > > > > > >> > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > >> > > > > > > > >> > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > >> > > > _cpuset); > > > > >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > > > >> > > > RTE_DEFINE_PER_LCORE(type, name) \ > > > > >> > > > lib/librte_eal/common/include/rte_eal.h: static > > > > >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > > > >> > > > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > > > >> > > > _rte_errno); > > > > >> > > > lib/librte_eal/common/eal_common_errno.c: static > > > > >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > >> > > > > > > > >> > > > > > > > >> > > > > > Thanks > > > > >> > > > > > Ravi > > > > >> > > > > > > > > > >> > > > > > Regards > > > > >> > > > > > > Neil > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > >> > > Its exported in the version map file: > > > > >> > > per_lcore__lcore_id; > > > > >> > > > > > > >> > > > > > > >> > Thanks Neil, I checked and both linux and bsd rte_eal_version.map > > have > > > > >> it. > > > > >> > I compared .map file between "changed code" and the original, > > they are > > > > >> same > > > > >> > for both linux and bsd. In fact you had ACK'd v4 version of this > > patch > > > > >> > series and no major changes after that. Please let me know if I > > missed > > > > >> > something. > > > > >> > > > > > >> I did, and I'm retracting that, because I didn't think to check the > > ABI > > > > >> compatibility on this. But I ran it throught the ABI checking > > script > > > > >> this and > > > > >> this error popped out. You should run it as well, its in the > > scripts > > > > >> directory. > > > > >> > > > > >> > > > > >> I see in your first patch you removed it and re-added it in the > > common > > > > >> section. > > > > >> But something about how its building is causing it to not show up > > as an > > > > >> exported > > > > >> symbol, which is problematic, as other applications are going to > > want > > > > >> access to > > > > >> it. > > > > >> > > > > >> It also possible that the ABI checker is throwing a false positive, > > but > > > > >> either > > > > >> way, it needs to be looked into prior to moving forward with this. > > > > >> > > > > >> > > > > > I did following things. > > > > > > > > > > Put a tag (v2.0.0-before-common-eal) before EAL common functions > > changes > > > > > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: remove > > unneeded > > > > > include) > > > > > > > > > > Put a tag (v2.0.0-common-eal) after EAL common functions changes for > > > > > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common > > functions in > > > > > eal_pci.c) > > > > > > > > > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > > > > > > > > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for > > librte_eal.so > > > > > shows removed symbols for "per_lcore__cpuset" > > > > > > > > > > v2.0.0-rc3 and v2.0.0-common-eal, html report for librte_eal.so shows > > > > > removed symbols for "per_lcore__cpuset" > > > > > > > > > > Removed symbol is different from what you have reported and in my > > case I > > > > > see it even before my commit. If you are interested I can unicast > > you html > > > > > report file. Please let me know how to proceed. > > > > > > > > > > > > > > > > > > I did some experiment and found some interesting things. I will take > > eal.c > > > > as an example > > > > > > > > eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c > > > > eal_common_proc_type.c and eal_common_app_usage.c. In > > linuxapp/eal/Makefile > > > > if I compile new files right after eal.c as shown below > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c > > > > ... > > > > > > > > validate-abi results matches baseline. Instead if i place new _common_ > > > > files in common area in linuxapp/eal/Makefile as shown below > > > > > > > > # from common dir > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c > > > > ... > > > > > > > > validate-abi reports problem in binary compatibility and source > > > > compatiblity > > > > > > > > eal_filesystem.h, librte_eal.so.1 > > > > [+] eal_parse_sysfs_value ( char const* filename, unsigned long* val ) > > > > @@ DPDK_2.0 (2) > > > > > > > > I believe files in common and linuxapp directory are compiled same way > > so > > > > not sure why placement in makefile makes difference. > > > > > > > > Could this be false-positive from validate-abi script?? > > > > > > > It could be, yes. Though I'm more inclined to think that perhaps in the > > new > > > version of the code we're not generating ithe same dwarf information out > > of it. > > > In fact for some reason, I've checked both the build before and after > > your > > > patch series, and the exported CFLAGS aren't getting passed to the build > > > properly, implying that we're not building all the code in the validator > > with > > > the -g flag, which the validator need to function properly. I'm looking > > into > > > that > > > Neil > > > > > > > > Found the problem, I was stupidly reading the report incorrectly. The > > problem > > regarding _lcore_id is a source compatibilty issue (because the symbol > > moved to > > a new location), which is irrelevant to us. Its not in any way a binary > > compat > > problem, which is what we care about. Sorry for the noise. > > > > I do still have a few concerns about some changed calling conventions with > > a few > > other functions, which I'll look into on monday. > > > > > Please let me know your inputs on changed calling conventions. Most of them > can be fixed by re-arranging moved code in _common_ files and order of > compilation. > If moving the order of compliation around fixes the problem, then I am reasonably convinced that it is, if not a false positive, a minor issue with the compilers dwarf information (The compiler just can't sanely change the location in which parameters are passed). If you make those changes, I'll ACK them, and look into whats going on with the calling conventions Thanks! Neil > Thanks, > Ravi > > Regards > > Neil > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-27 13:44 ` Neil Horman @ 2015-04-27 22:39 ` Ravi Kerur 2015-04-28 19:35 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-27 22:39 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Mon, Apr 27, 2015 at 6:44 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > On Sat, Apr 25, 2015 at 05:09:01PM -0700, Ravi Kerur wrote: > > On Sat, Apr 25, 2015 at 6:02 AM, Neil Horman <nhorman@tuxdriver.com> > wrote: > > > > > On Sat, Apr 25, 2015 at 08:32:42AM -0400, Neil Horman wrote: > > > > On Fri, Apr 24, 2015 at 06:45:06PM -0700, Ravi Kerur wrote: > > > > > On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> > wrote: > > > > > > > > > > > > > > > > > > > > > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman < > nhorman@tuxdriver.com > > > > > > > > > > wrote: > > > > > > > > > > > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > > > > > >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman < > > > nhorman@tuxdriver.com> > > > > > >> wrote: > > > > > >> > > > > > > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > > > > > >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman < > > > nhorman@tuxdriver.com > > > > > >> > > > > > > >> > > wrote: > > > > > >> > > > > > > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur > wrote: > > > > > >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > > > > > >> nhorman@tuxdriver.com> > > > > > >> > > > > wrote: > > > > > >> > > > > > > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur > > > wrote: > > > > > >> > > > > > > > Changes in v7 > > > > > >> > > > > > > > Remove _setname_ pthread calls. > > > > > >> > > > > > > > Use rte_gettid() API in RTE_LOG to print > thread_id. > > > > > >> > > > > > > > > > > > > >> > > > > > > > Changes in v6 > > > > > >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from > eal_common_thread.c > > > file. > > > > > >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for > > > Linux/FreeBSD > > > > > >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG > when > > > > > >> available. > > > > > >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to print > > > thread_id. > > > > > >> > > > > > > > > > > > > >> > > > > > > > Changes in v5 > > > > > >> > > > > > > > Rebase to latest code. > > > > > >> > > > > > > > > > > > > >> > > > > > > > Changes in v4 > > > > > >> > > > > > > > None > > > > > >> > > > > > > > > > > > > >> > > > > > > > Changes in v3 > > > > > >> > > > > > > > Changed subject to be more explicit on file name > > > inclusion. > > > > > >> > > > > > > > > > > > > >> > > > > > > > Changes in v2 > > > > > >> > > > > > > > None > > > > > >> > > > > > > > > > > > > >> > > > > > > > Changes in v1 > > > > > >> > > > > > > > eal_thread.c has minor differences between Linux > and > > > BSD, > > > > > >> move > > > > > >> > > > > > > > entire file into common directory. > > > > > >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor > > > > > >> differences. > > > > > >> > > > > > > > Rename eal_thread.c to eal_common_thread.c > > > > > >> > > > > > > > Makefile changes to reflect file move and name > change. > > > > > >> > > > > > > > Fix checkpatch warnings. > > > > > >> > > > > > > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > > >> > > > > > > > --- > > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 > +- > > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > > > >> > > > > > > ------------------------------ > > > > > >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > > > >> > > > > > > ++++++++++++++++++++++++++++- > > > > > >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > > > >> > > > > > > +----------------------------- > > > > > >> > > > > > > > 4 files changed, 148 insertions(+), 305 > deletions(-) > > > > > >> > > > > > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > > >> > > > > > > > index 2357cfa..55971b9 100644 > > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > > >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := > > > -D_GNU_SOURCE > > > > > >> > > > > > > > # workaround for a gcc bug with noreturn > attribute > > > > > >> > > > > > > > # > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > > >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > > >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > > >> > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > > > >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > > >> > > > > > > > endif > > > > > >> > > > > > > > > > > > > >> > > > > > > > diff --git > a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > >> > > > > > > > index 9a03437..5714b8f 100644 > > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > >> > > > > > > > @@ -35,163 +35,11 @@ > > > > > >> > > > > > > > #include <stdio.h> > > > > > >> > > > > > > > #include <stdlib.h> > > > > > >> > > > > > > > #include <stdint.h> > > > > > >> > > > > > > > -#include <unistd.h> > > > > > >> > > > > > > > -#include <sched.h> > > > > > >> > > > > > > > -#include <pthread_np.h> > > > > > >> > > > > > > > -#include <sys/queue.h> > > > > > >> > > > > > > > #include <sys/thr.h> > > > > > >> > > > > > > > > > > > > >> > > > > > > > -#include <rte_debug.h> > > > > > >> > > > > > > > -#include <rte_atomic.h> > > > > > >> > > > > > > > -#include <rte_launch.h> > > > > > >> > > > > > > > -#include <rte_log.h> > > > > > >> > > > > > > > -#include <rte_memory.h> > > > > > >> > > > > > > > -#include <rte_memzone.h> > > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > > >> > > > > > > > -#include <rte_eal.h> > > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > > >> > > > > > > > -#include <rte_lcore.h> > > > > > >> > > > > > > > - > > > > > >> > > > > > > > #include "eal_private.h" > > > > > >> > > > > > > > #include "eal_thread.h" > > > > > >> > > > > > > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = > > > LCORE_ID_ANY; > > > > > >> > > > > > > NAK, these are exported symbols, you can't remove > them > > > without > > > > > >> > > going > > > > > >> > > > > > > through the > > > > > >> > > > > > > deprecation process. > > > > > >> > > > > > > > > > > > >> > > > > > > > > > > > >> > > > > > They are not removed/deleted, they are moved from > > > eal_thread.c > > > > > >> to > > > > > >> > > > > > eal_common_thread.c file since it is common to both > Linux > > > and > > > > > >> BSD. > > > > > >> > > > > > > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its > showing > > > up as > > > > > >> > > removed > > > > > >> > > > > on the > > > > > >> > > > > ABI checker utility. > > > > > >> > > > > > > > > > >> > > > > Neil > > > > > >> > > > > > > > > > >> > > > > > > > > >> > > > Can you please show me in the current code where it is > being > > > > > >> exported? I > > > > > >> > > > have only moved definitions to _common_ files, not sure > why it > > > > > >> should be > > > > > >> > > > exported now. I searched in the current code for > > > > > >> RTE_DEFINE_PER_LCORE > > > > > >> > > > > > > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE > * > > > > > >> > > > app/test/test_per_lcore.c:static > > > RTE_DEFINE_PER_LCORE(unsigned, > > > > > >> test) = > > > > > >> > > > 0x12345678; > > > > > >> > > > > > > > > >> > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > > >> > > > > > > > > >> > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > >> > > > > > > > > >> > > > > > > > >> > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > > >> > > > _cpuset); > > > > > >> > > > > > > > > >> > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > > >> > > > > > > > > >> > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > >> > > > > > > > > >> > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > > >> > > > _cpuset); > > > > > >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > > > > >> > > > RTE_DEFINE_PER_LCORE(type, name) \ > > > > > >> > > > lib/librte_eal/common/include/rte_eal.h: static > > > > > >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > > > > >> > > > > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > > > > >> > > > _rte_errno); > > > > > >> > > > lib/librte_eal/common/eal_common_errno.c: static > > > > > >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > > >> > > > > > > > > >> > > > > > > > > >> > > > > > Thanks > > > > > >> > > > > > Ravi > > > > > >> > > > > > > > > > > >> > > > > > Regards > > > > > >> > > > > > > Neil > > > > > >> > > > > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > > >> > > Its exported in the version map file: > > > > > >> > > per_lcore__lcore_id; > > > > > >> > > > > > > > >> > > > > > > > >> > Thanks Neil, I checked and both linux and bsd > rte_eal_version.map > > > have > > > > > >> it. > > > > > >> > I compared .map file between "changed code" and the original, > > > they are > > > > > >> same > > > > > >> > for both linux and bsd. In fact you had ACK'd v4 version of > this > > > patch > > > > > >> > series and no major changes after that. Please let me know if > I > > > missed > > > > > >> > something. > > > > > >> > > > > > > >> I did, and I'm retracting that, because I didn't think to check > the > > > ABI > > > > > >> compatibility on this. But I ran it throught the ABI checking > > > script > > > > > >> this and > > > > > >> this error popped out. You should run it as well, its in the > > > scripts > > > > > >> directory. > > > > > >> > > > > > >> > > > > > >> I see in your first patch you removed it and re-added it in the > > > common > > > > > >> section. > > > > > >> But something about how its building is causing it to not show > up > > > as an > > > > > >> exported > > > > > >> symbol, which is problematic, as other applications are going to > > > want > > > > > >> access to > > > > > >> it. > > > > > >> > > > > > >> It also possible that the ABI checker is throwing a false > positive, > > > but > > > > > >> either > > > > > >> way, it needs to be looked into prior to moving forward with > this. > > > > > >> > > > > > >> > > > > > > I did following things. > > > > > > > > > > > > Put a tag (v2.0.0-before-common-eal) before EAL common functions > > > changes > > > > > > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: remove > > > unneeded > > > > > > include) > > > > > > > > > > > > Put a tag (v2.0.0-common-eal) after EAL common functions changes > for > > > > > > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common > > > functions in > > > > > > eal_pci.c) > > > > > > > > > > > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > > > > > > > > > > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for > > > librte_eal.so > > > > > > shows removed symbols for "per_lcore__cpuset" > > > > > > > > > > > > v2.0.0-rc3 and v2.0.0-common-eal, html report for librte_eal.so > shows > > > > > > removed symbols for "per_lcore__cpuset" > > > > > > > > > > > > Removed symbol is different from what you have reported and in my > > > case I > > > > > > see it even before my commit. If you are interested I can unicast > > > you html > > > > > > report file. Please let me know how to proceed. > > > > > > > > > > > > > > > > > > > > > > I did some experiment and found some interesting things. I will > take > > > eal.c > > > > > as an example > > > > > > > > > > eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c > > > > > eal_common_proc_type.c and eal_common_app_usage.c. In > > > linuxapp/eal/Makefile > > > > > if I compile new files right after eal.c as shown below > > > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c > > > > > ... > > > > > > > > > > validate-abi results matches baseline. Instead if i place new > _common_ > > > > > files in common area in linuxapp/eal/Makefile as shown below > > > > > > > > > > # from common dir > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c > > > > > ... > > > > > > > > > > validate-abi reports problem in binary compatibility and source > > > > > compatiblity > > > > > > > > > > eal_filesystem.h, librte_eal.so.1 > > > > > [+] eal_parse_sysfs_value ( char const* filename, unsigned long* > val ) > > > > > @@ DPDK_2.0 (2) > > > > > > > > > > I believe files in common and linuxapp directory are compiled same > way > > > so > > > > > not sure why placement in makefile makes difference. > > > > > > > > > > Could this be false-positive from validate-abi script?? > > > > > > > > > It could be, yes. Though I'm more inclined to think that perhaps in > the > > > new > > > > version of the code we're not generating ithe same dwarf information > out > > > of it. > > > > In fact for some reason, I've checked both the build before and after > > > your > > > > patch series, and the exported CFLAGS aren't getting passed to the > build > > > > properly, implying that we're not building all the code in the > validator > > > with > > > > the -g flag, which the validator need to function properly. I'm > looking > > > into > > > > that > > > > Neil > > > > > > > > > > > Found the problem, I was stupidly reading the report incorrectly. The > > > problem > > > regarding _lcore_id is a source compatibilty issue (because the symbol > > > moved to > > > a new location), which is irrelevant to us. Its not in any way a > binary > > > compat > > > problem, which is what we care about. Sorry for the noise. > > > > > > I do still have a few concerns about some changed calling conventions > with > > > a few > > > other functions, which I'll look into on monday. > > > > > > > > Please let me know your inputs on changed calling conventions. Most of > them > > can be fixed by re-arranging moved code in _common_ files and order of > > compilation. > > > If moving the order of compliation around fixes the problem, then I am > reasonably convinced that it is, if not a false positive, a minor issue > with the > compilers dwarf information (The compiler just can't sanely change the > location > in which parameters are passed). If you make those changes, I'll ACK > them, and > look into whats going on with the calling conventions > Issues like the one shown below are taken care by reordering the code compilation. eal_parse_sysfs_value ( char const* filename, unsigned long* val ) Change The parameter filename became passed on stack instead of rdi register Effect Violation of the calling convention. This may result in crash or incorrect behavior of applications. Last one that is left out is in rte_thread_set_affinity ( rte_cpuset_t* p1 ) Change The parameter *p1* became passed in *rdi* register instead of stack. Effect Violation of the calling convention. This may result in crash or incorrect behavior of applications. After checking abi-0.99.pdf (x86-64.org) looks like for "rte_thread_set_affinity" new code is doing the right thing by passing the parameter in "rdi" register since pointer is classified as "integer_class". Nothing needs to be fixed here. After you confirm that warning can be ignored I will work on sending new revision. Thanks, Ravi > Thanks! > Neil > > > Thanks, > > Ravi > > > > Regards > > > Neil > > > > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-27 22:39 ` Ravi Kerur @ 2015-04-28 19:35 ` Neil Horman 2015-04-28 23:52 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-28 19:35 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Mon, Apr 27, 2015 at 03:39:41PM -0700, Ravi Kerur wrote: > On Mon, Apr 27, 2015 at 6:44 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > > > On Sat, Apr 25, 2015 at 05:09:01PM -0700, Ravi Kerur wrote: > > > On Sat, Apr 25, 2015 at 6:02 AM, Neil Horman <nhorman@tuxdriver.com> > > wrote: > > > > > > > On Sat, Apr 25, 2015 at 08:32:42AM -0400, Neil Horman wrote: > > > > > On Fri, Apr 24, 2015 at 06:45:06PM -0700, Ravi Kerur wrote: > > > > > > On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman < > > nhorman@tuxdriver.com > > > > > > > > > > > > wrote: > > > > > > > > > > > > > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > > > > > > >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman < > > > > nhorman@tuxdriver.com> > > > > > > >> wrote: > > > > > > >> > > > > > > > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur wrote: > > > > > > >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman < > > > > nhorman@tuxdriver.com > > > > > > >> > > > > > > > >> > > wrote: > > > > > > >> > > > > > > > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur > > wrote: > > > > > > >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > > > > > > >> nhorman@tuxdriver.com> > > > > > > >> > > > > wrote: > > > > > > >> > > > > > > > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi Kerur > > > > wrote: > > > > > > >> > > > > > > > Changes in v7 > > > > > > >> > > > > > > > Remove _setname_ pthread calls. > > > > > > >> > > > > > > > Use rte_gettid() API in RTE_LOG to print > > thread_id. > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > Changes in v6 > > > > > > >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from > > eal_common_thread.c > > > > file. > > > > > > >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for > > > > Linux/FreeBSD > > > > > > >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG > > when > > > > > > >> available. > > > > > > >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to print > > > > thread_id. > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > Changes in v5 > > > > > > >> > > > > > > > Rebase to latest code. > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > Changes in v4 > > > > > > >> > > > > > > > None > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > Changes in v3 > > > > > > >> > > > > > > > Changed subject to be more explicit on file name > > > > inclusion. > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > Changes in v2 > > > > > > >> > > > > > > > None > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > Changes in v1 > > > > > > >> > > > > > > > eal_thread.c has minor differences between Linux > > and > > > > BSD, > > > > > > >> move > > > > > > >> > > > > > > > entire file into common directory. > > > > > > >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on minor > > > > > > >> differences. > > > > > > >> > > > > > > > Rename eal_thread.c to eal_common_thread.c > > > > > > >> > > > > > > > Makefile changes to reflect file move and name > > change. > > > > > > >> > > > > > > > Fix checkpatch warnings. > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > > > >> > > > > > > > --- > > > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | 2 > > +- > > > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | 152 > > > > > > >> > > > > > > ------------------------------ > > > > > > >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | 147 > > > > > > >> > > > > > > ++++++++++++++++++++++++++++- > > > > > > >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | 152 > > > > > > >> > > > > > > +----------------------------- > > > > > > >> > > > > > > > 4 files changed, 148 insertions(+), 305 > > deletions(-) > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > diff --git a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > >> > > > > > > > index 2357cfa..55971b9 100644 > > > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := > > > > -D_GNU_SOURCE > > > > > > >> > > > > > > > # workaround for a gcc bug with noreturn > > attribute > > > > > > >> > > > > > > > # > > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > > > >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > > > >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > > > >> > > > > > > > +CFLAGS_eal_common_thread.o += -Wno-return-type > > > > > > >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > > > >> > > > > > > > endif > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > diff --git > > a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > >> > > > > > > > index 9a03437..5714b8f 100644 > > > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > >> > > > > > > > @@ -35,163 +35,11 @@ > > > > > > >> > > > > > > > #include <stdio.h> > > > > > > >> > > > > > > > #include <stdlib.h> > > > > > > >> > > > > > > > #include <stdint.h> > > > > > > >> > > > > > > > -#include <unistd.h> > > > > > > >> > > > > > > > -#include <sched.h> > > > > > > >> > > > > > > > -#include <pthread_np.h> > > > > > > >> > > > > > > > -#include <sys/queue.h> > > > > > > >> > > > > > > > #include <sys/thr.h> > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > -#include <rte_debug.h> > > > > > > >> > > > > > > > -#include <rte_atomic.h> > > > > > > >> > > > > > > > -#include <rte_launch.h> > > > > > > >> > > > > > > > -#include <rte_log.h> > > > > > > >> > > > > > > > -#include <rte_memory.h> > > > > > > >> > > > > > > > -#include <rte_memzone.h> > > > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > > > >> > > > > > > > -#include <rte_eal.h> > > > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > > > >> > > > > > > > -#include <rte_lcore.h> > > > > > > >> > > > > > > > - > > > > > > >> > > > > > > > #include "eal_private.h" > > > > > > >> > > > > > > > #include "eal_thread.h" > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = > > > > LCORE_ID_ANY; > > > > > > >> > > > > > > NAK, these are exported symbols, you can't remove > > them > > > > without > > > > > > >> > > going > > > > > > >> > > > > > > through the > > > > > > >> > > > > > > deprecation process. > > > > > > >> > > > > > > > > > > > > >> > > > > > > > > > > > > >> > > > > > They are not removed/deleted, they are moved from > > > > eal_thread.c > > > > > > >> to > > > > > > >> > > > > > eal_common_thread.c file since it is common to both > > Linux > > > > and > > > > > > >> BSD. > > > > > > >> > > > > > > > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its > > showing > > > > up as > > > > > > >> > > removed > > > > > > >> > > > > on the > > > > > > >> > > > > ABI checker utility. > > > > > > >> > > > > > > > > > > >> > > > > Neil > > > > > > >> > > > > > > > > > > >> > > > > > > > > > >> > > > Can you please show me in the current code where it is > > being > > > > > > >> exported? I > > > > > > >> > > > have only moved definitions to _common_ files, not sure > > why it > > > > > > >> should be > > > > > > >> > > > exported now. I searched in the current code for > > > > > > >> RTE_DEFINE_PER_LCORE > > > > > > >> > > > > > > > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir RTE_DEFINE_PER_LCORE > > * > > > > > > >> > > > app/test/test_per_lcore.c:static > > > > RTE_DEFINE_PER_LCORE(unsigned, > > > > > > >> test) = > > > > > > >> > > > 0x12345678; > > > > > > >> > > > > > > > > > >> > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > > > >> > > > > > > > > > >> > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > > >> > > > > > > > > > >> > > > > > > > > >> > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > > > >> > > > _cpuset); > > > > > > >> > > > > > > > > > >> > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > > > >> > > > > > > > > > >> > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > > >> > > > > > > > > > >> > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > > > >> > > > _cpuset); > > > > > > >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > > > > > >> > > > RTE_DEFINE_PER_LCORE(type, name) \ > > > > > > >> > > > lib/librte_eal/common/include/rte_eal.h: static > > > > > > >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > > > > > >> > > > > > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > > > > > >> > > > _rte_errno); > > > > > > >> > > > lib/librte_eal/common/eal_common_errno.c: static > > > > > > >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > > > >> > > > > > > > > > >> > > > > > > > > > >> > > > > > Thanks > > > > > > >> > > > > > Ravi > > > > > > >> > > > > > > > > > > > >> > > > > > Regards > > > > > > >> > > > > > > Neil > > > > > > >> > > > > > > > > > > > > >> > > > > > > > > > > > > >> > > > > > > > > > > >> > > Its exported in the version map file: > > > > > > >> > > per_lcore__lcore_id; > > > > > > >> > > > > > > > > >> > > > > > > > > >> > Thanks Neil, I checked and both linux and bsd > > rte_eal_version.map > > > > have > > > > > > >> it. > > > > > > >> > I compared .map file between "changed code" and the original, > > > > they are > > > > > > >> same > > > > > > >> > for both linux and bsd. In fact you had ACK'd v4 version of > > this > > > > patch > > > > > > >> > series and no major changes after that. Please let me know if > > I > > > > missed > > > > > > >> > something. > > > > > > >> > > > > > > > >> I did, and I'm retracting that, because I didn't think to check > > the > > > > ABI > > > > > > >> compatibility on this. But I ran it throught the ABI checking > > > > script > > > > > > >> this and > > > > > > >> this error popped out. You should run it as well, its in the > > > > scripts > > > > > > >> directory. > > > > > > >> > > > > > > >> > > > > > > >> I see in your first patch you removed it and re-added it in the > > > > common > > > > > > >> section. > > > > > > >> But something about how its building is causing it to not show > > up > > > > as an > > > > > > >> exported > > > > > > >> symbol, which is problematic, as other applications are going to > > > > want > > > > > > >> access to > > > > > > >> it. > > > > > > >> > > > > > > >> It also possible that the ABI checker is throwing a false > > positive, > > > > but > > > > > > >> either > > > > > > >> way, it needs to be looked into prior to moving forward with > > this. > > > > > > >> > > > > > > >> > > > > > > > I did following things. > > > > > > > > > > > > > > Put a tag (v2.0.0-before-common-eal) before EAL common functions > > > > changes > > > > > > > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: remove > > > > unneeded > > > > > > > include) > > > > > > > > > > > > > > Put a tag (v2.0.0-common-eal) after EAL common functions changes > > for > > > > > > > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common > > > > functions in > > > > > > > eal_pci.c) > > > > > > > > > > > > > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > > > > > > > > > > > > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for > > > > librte_eal.so > > > > > > > shows removed symbols for "per_lcore__cpuset" > > > > > > > > > > > > > > v2.0.0-rc3 and v2.0.0-common-eal, html report for librte_eal.so > > shows > > > > > > > removed symbols for "per_lcore__cpuset" > > > > > > > > > > > > > > Removed symbol is different from what you have reported and in my > > > > case I > > > > > > > see it even before my commit. If you are interested I can unicast > > > > you html > > > > > > > report file. Please let me know how to proceed. > > > > > > > > > > > > > > > > > > > > > > > > > > I did some experiment and found some interesting things. I will > > take > > > > eal.c > > > > > > as an example > > > > > > > > > > > > eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c > > > > > > eal_common_proc_type.c and eal_common_app_usage.c. In > > > > linuxapp/eal/Makefile > > > > > > if I compile new files right after eal.c as shown below > > > > > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_proc_type.c > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_app_usage.c > > > > > > ... > > > > > > > > > > > > validate-abi results matches baseline. Instead if i place new > > _common_ > > > > > > files in common area in linuxapp/eal/Makefile as shown below > > > > > > > > > > > > # from common dir > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c > > > > > > ... > > > > > > > > > > > > validate-abi reports problem in binary compatibility and source > > > > > > compatiblity > > > > > > > > > > > > eal_filesystem.h, librte_eal.so.1 > > > > > > [+] eal_parse_sysfs_value ( char const* filename, unsigned long* > > val ) > > > > > > @@ DPDK_2.0 (2) > > > > > > > > > > > > I believe files in common and linuxapp directory are compiled same > > way > > > > so > > > > > > not sure why placement in makefile makes difference. > > > > > > > > > > > > Could this be false-positive from validate-abi script?? > > > > > > > > > > > It could be, yes. Though I'm more inclined to think that perhaps in > > the > > > > new > > > > > version of the code we're not generating ithe same dwarf information > > out > > > > of it. > > > > > In fact for some reason, I've checked both the build before and after > > > > your > > > > > patch series, and the exported CFLAGS aren't getting passed to the > > build > > > > > properly, implying that we're not building all the code in the > > validator > > > > with > > > > > the -g flag, which the validator need to function properly. I'm > > looking > > > > into > > > > > that > > > > > Neil > > > > > > > > > > > > > > Found the problem, I was stupidly reading the report incorrectly. The > > > > problem > > > > regarding _lcore_id is a source compatibilty issue (because the symbol > > > > moved to > > > > a new location), which is irrelevant to us. Its not in any way a > > binary > > > > compat > > > > problem, which is what we care about. Sorry for the noise. > > > > > > > > I do still have a few concerns about some changed calling conventions > > with > > > > a few > > > > other functions, which I'll look into on monday. > > > > > > > > > > > Please let me know your inputs on changed calling conventions. Most of > > them > > > can be fixed by re-arranging moved code in _common_ files and order of > > > compilation. > > > > > If moving the order of compliation around fixes the problem, then I am > > reasonably convinced that it is, if not a false positive, a minor issue > > with the > > compilers dwarf information (The compiler just can't sanely change the > > location > > in which parameters are passed). If you make those changes, I'll ACK > > them, and > > look into whats going on with the calling conventions > > > > Issues like the one shown below are taken care by reordering the code > compilation. > > eal_parse_sysfs_value ( char const* filename, unsigned long* val ) > > Change > The parameter filename became passed on stack instead of rdi register > > Effect > Violation of the calling convention. This may result in crash or incorrect > behavior of applications. > > Last one that is left out is in > > rte_thread_set_affinity ( rte_cpuset_t* p1 ) > > Change > The parameter *p1* became passed in *rdi* register instead of stack. > > Effect > Violation of the calling convention. This may result in crash or incorrect > behavior of applications. > > After checking abi-0.99.pdf (x86-64.org) looks like for > "rte_thread_set_affinity" new code is doing the right thing by passing the > parameter in "rdi" register since pointer is classified as "integer_class". > Nothing needs to be fixed here. After you confirm that warning can be > ignored I will work on sending new revision. > ACK then, send the new revision, this appears to be a false positive. Thanks for taking the time to confirm. Best Neil > Thanks, > Ravi > > > > Thanks! > > Neil > > > > > Thanks, > > > Ravi > > > > > > Regards > > > > Neil > > > > > > > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-28 19:35 ` Neil Horman @ 2015-04-28 23:52 ` Ravi Kerur 2015-04-29 10:04 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-28 23:52 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Tue, Apr 28, 2015 at 12:35 PM, Neil Horman <nhorman@tuxdriver.com> wrote: > On Mon, Apr 27, 2015 at 03:39:41PM -0700, Ravi Kerur wrote: > > On Mon, Apr 27, 2015 at 6:44 AM, Neil Horman <nhorman@tuxdriver.com> > wrote: > > > > > On Sat, Apr 25, 2015 at 05:09:01PM -0700, Ravi Kerur wrote: > > > > On Sat, Apr 25, 2015 at 6:02 AM, Neil Horman <nhorman@tuxdriver.com> > > > wrote: > > > > > > > > > On Sat, Apr 25, 2015 at 08:32:42AM -0400, Neil Horman wrote: > > > > > > On Fri, Apr 24, 2015 at 06:45:06PM -0700, Ravi Kerur wrote: > > > > > > > On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman < > > > nhorman@tuxdriver.com > > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > > > > > > > >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman < > > > > > nhorman@tuxdriver.com> > > > > > > > >> wrote: > > > > > > > >> > > > > > > > > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur > wrote: > > > > > > > >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman < > > > > > nhorman@tuxdriver.com > > > > > > > >> > > > > > > > > >> > > wrote: > > > > > > > >> > > > > > > > > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur > > > wrote: > > > > > > > >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > > > > > > > >> nhorman@tuxdriver.com> > > > > > > > >> > > > > wrote: > > > > > > > >> > > > > > > > > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi > Kerur > > > > > wrote: > > > > > > > >> > > > > > > > Changes in v7 > > > > > > > >> > > > > > > > Remove _setname_ pthread calls. > > > > > > > >> > > > > > > > Use rte_gettid() API in RTE_LOG to print > > > thread_id. > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > Changes in v6 > > > > > > > >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from > > > eal_common_thread.c > > > > > file. > > > > > > > >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for > > > > > Linux/FreeBSD > > > > > > > >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG > > > when > > > > > > > >> available. > > > > > > > >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to > print > > > > > thread_id. > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > Changes in v5 > > > > > > > >> > > > > > > > Rebase to latest code. > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > Changes in v4 > > > > > > > >> > > > > > > > None > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > Changes in v3 > > > > > > > >> > > > > > > > Changed subject to be more explicit on file > name > > > > > inclusion. > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > Changes in v2 > > > > > > > >> > > > > > > > None > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > Changes in v1 > > > > > > > >> > > > > > > > eal_thread.c has minor differences between > Linux > > > and > > > > > BSD, > > > > > > > >> move > > > > > > > >> > > > > > > > entire file into common directory. > > > > > > > >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on > minor > > > > > > > >> differences. > > > > > > > >> > > > > > > > Rename eal_thread.c to eal_common_thread.c > > > > > > > >> > > > > > > > Makefile changes to reflect file move and name > > > change. > > > > > > > >> > > > > > > > Fix checkpatch warnings. > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > > > > >> > > > > > > > --- > > > > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | > 2 > > > +- > > > > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | > 152 > > > > > > > >> > > > > > > ------------------------------ > > > > > > > >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | > 147 > > > > > > > >> > > > > > > ++++++++++++++++++++++++++++- > > > > > > > >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | > 152 > > > > > > > >> > > > > > > +----------------------------- > > > > > > > >> > > > > > > > 4 files changed, 148 insertions(+), 305 > > > deletions(-) > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > diff --git > a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > >> > > > > > > > index 2357cfa..55971b9 100644 > > > > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := > > > > > -D_GNU_SOURCE > > > > > > > >> > > > > > > > # workaround for a gcc bug with noreturn > > > attribute > > > > > > > >> > > > > > > > # > > > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > > > > >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > > > > >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > > > > >> > > > > > > > +CFLAGS_eal_common_thread.o += > -Wno-return-type > > > > > > > >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > > > > >> > > > > > > > endif > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > diff --git > > > a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > >> > > > > > > > index 9a03437..5714b8f 100644 > > > > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > >> > > > > > > > @@ -35,163 +35,11 @@ > > > > > > > >> > > > > > > > #include <stdio.h> > > > > > > > >> > > > > > > > #include <stdlib.h> > > > > > > > >> > > > > > > > #include <stdint.h> > > > > > > > >> > > > > > > > -#include <unistd.h> > > > > > > > >> > > > > > > > -#include <sched.h> > > > > > > > >> > > > > > > > -#include <pthread_np.h> > > > > > > > >> > > > > > > > -#include <sys/queue.h> > > > > > > > >> > > > > > > > #include <sys/thr.h> > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > -#include <rte_debug.h> > > > > > > > >> > > > > > > > -#include <rte_atomic.h> > > > > > > > >> > > > > > > > -#include <rte_launch.h> > > > > > > > >> > > > > > > > -#include <rte_log.h> > > > > > > > >> > > > > > > > -#include <rte_memory.h> > > > > > > > >> > > > > > > > -#include <rte_memzone.h> > > > > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > > > > >> > > > > > > > -#include <rte_eal.h> > > > > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > > > > >> > > > > > > > -#include <rte_lcore.h> > > > > > > > >> > > > > > > > - > > > > > > > >> > > > > > > > #include "eal_private.h" > > > > > > > >> > > > > > > > #include "eal_thread.h" > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = > > > > > LCORE_ID_ANY; > > > > > > > >> > > > > > > NAK, these are exported symbols, you can't > remove > > > them > > > > > without > > > > > > > >> > > going > > > > > > > >> > > > > > > through the > > > > > > > >> > > > > > > deprecation process. > > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > > > > > > > >> > > > > > They are not removed/deleted, they are moved from > > > > > eal_thread.c > > > > > > > >> to > > > > > > > >> > > > > > eal_common_thread.c file since it is common to > both > > > Linux > > > > > and > > > > > > > >> BSD. > > > > > > > >> > > > > > > > > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its > > > showing > > > > > up as > > > > > > > >> > > removed > > > > > > > >> > > > > on the > > > > > > > >> > > > > ABI checker utility. > > > > > > > >> > > > > > > > > > > > >> > > > > Neil > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > > > >> > > > Can you please show me in the current code where it is > > > being > > > > > > > >> exported? I > > > > > > > >> > > > have only moved definitions to _common_ files, not > sure > > > why it > > > > > > > >> should be > > > > > > > >> > > > exported now. I searched in the current code for > > > > > > > >> RTE_DEFINE_PER_LCORE > > > > > > > >> > > > > > > > > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir > RTE_DEFINE_PER_LCORE > > > * > > > > > > > >> > > > app/test/test_per_lcore.c:static > > > > > RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > >> test) = > > > > > > > >> > > > 0x12345678; > > > > > > > >> > > > > > > > > > > >> > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > > > > >> > > > > > > > > > > >> > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > >> > > > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > > > > >> > > > _cpuset); > > > > > > > >> > > > > > > > > > > >> > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > > > > >> > > > > > > > > > > >> > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > > > > >> > > > _cpuset); > > > > > > > >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > > > > > > >> > > > RTE_DEFINE_PER_LCORE(type, name) \ > > > > > > > >> > > > lib/librte_eal/common/include/rte_eal.h: static > > > > > > > >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > > > > > > >> > > > > > > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > > > > > > >> > > > _rte_errno); > > > > > > > >> > > > lib/librte_eal/common/eal_common_errno.c: static > > > > > > > >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > > > > > Thanks > > > > > > > >> > > > > > Ravi > > > > > > > >> > > > > > > > > > > > > >> > > > > > Regards > > > > > > > >> > > > > > > Neil > > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > > > > > > > >> > > > > > > > > > > > >> > > Its exported in the version map file: > > > > > > > >> > > per_lcore__lcore_id; > > > > > > > >> > > > > > > > > > >> > > > > > > > > > >> > Thanks Neil, I checked and both linux and bsd > > > rte_eal_version.map > > > > > have > > > > > > > >> it. > > > > > > > >> > I compared .map file between "changed code" and the > original, > > > > > they are > > > > > > > >> same > > > > > > > >> > for both linux and bsd. In fact you had ACK'd v4 version > of > > > this > > > > > patch > > > > > > > >> > series and no major changes after that. Please let me > know if > > > I > > > > > missed > > > > > > > >> > something. > > > > > > > >> > > > > > > > > >> I did, and I'm retracting that, because I didn't think to > check > > > the > > > > > ABI > > > > > > > >> compatibility on this. But I ran it throught the ABI > checking > > > > > script > > > > > > > >> this and > > > > > > > >> this error popped out. You should run it as well, its in > the > > > > > scripts > > > > > > > >> directory. > > > > > > > >> > > > > > > > >> > > > > > > > >> I see in your first patch you removed it and re-added it in > the > > > > > common > > > > > > > >> section. > > > > > > > >> But something about how its building is causing it to not > show > > > up > > > > > as an > > > > > > > >> exported > > > > > > > >> symbol, which is problematic, as other applications are > going to > > > > > want > > > > > > > >> access to > > > > > > > >> it. > > > > > > > >> > > > > > > > >> It also possible that the ABI checker is throwing a false > > > positive, > > > > > but > > > > > > > >> either > > > > > > > >> way, it needs to be looked into prior to moving forward with > > > this. > > > > > > > >> > > > > > > > >> > > > > > > > > I did following things. > > > > > > > > > > > > > > > > Put a tag (v2.0.0-before-common-eal) before EAL common > functions > > > > > changes > > > > > > > > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: > remove > > > > > unneeded > > > > > > > > include) > > > > > > > > > > > > > > > > Put a tag (v2.0.0-common-eal) after EAL common functions > changes > > > for > > > > > > > > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common > > > > > functions in > > > > > > > > eal_pci.c) > > > > > > > > > > > > > > > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > > > > > > > > > > > > > > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for > > > > > librte_eal.so > > > > > > > > shows removed symbols for "per_lcore__cpuset" > > > > > > > > > > > > > > > > v2.0.0-rc3 and v2.0.0-common-eal, html report for > librte_eal.so > > > shows > > > > > > > > removed symbols for "per_lcore__cpuset" > > > > > > > > > > > > > > > > Removed symbol is different from what you have reported and > in my > > > > > case I > > > > > > > > see it even before my commit. If you are interested I can > unicast > > > > > you html > > > > > > > > report file. Please let me know how to proceed. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > I did some experiment and found some interesting things. I > will > > > take > > > > > eal.c > > > > > > > as an example > > > > > > > > > > > > > > eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c > > > > > > > eal_common_proc_type.c and eal_common_app_usage.c. In > > > > > linuxapp/eal/Makefile > > > > > > > if I compile new files right after eal.c as shown below > > > > > > > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += > eal_common_proc_type.c > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += > eal_common_app_usage.c > > > > > > > ... > > > > > > > > > > > > > > validate-abi results matches baseline. Instead if i place new > > > _common_ > > > > > > > files in common area in linuxapp/eal/Makefile as shown below > > > > > > > > > > > > > > # from common dir > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c > > > > > > > ... > > > > > > > > > > > > > > validate-abi reports problem in binary compatibility and source > > > > > > > compatiblity > > > > > > > > > > > > > > eal_filesystem.h, librte_eal.so.1 > > > > > > > [+] eal_parse_sysfs_value ( char const* filename, unsigned > long* > > > val ) > > > > > > > @@ DPDK_2.0 (2) > > > > > > > > > > > > > > I believe files in common and linuxapp directory are compiled > same > > > way > > > > > so > > > > > > > not sure why placement in makefile makes difference. > > > > > > > > > > > > > > Could this be false-positive from validate-abi script?? > > > > > > > > > > > > > It could be, yes. Though I'm more inclined to think that > perhaps in > > > the > > > > > new > > > > > > version of the code we're not generating ithe same dwarf > information > > > out > > > > > of it. > > > > > > In fact for some reason, I've checked both the build before and > after > > > > > your > > > > > > patch series, and the exported CFLAGS aren't getting passed to > the > > > build > > > > > > properly, implying that we're not building all the code in the > > > validator > > > > > with > > > > > > the -g flag, which the validator need to function properly. I'm > > > looking > > > > > into > > > > > > that > > > > > > Neil > > > > > > > > > > > > > > > > > Found the problem, I was stupidly reading the report incorrectly. > The > > > > > problem > > > > > regarding _lcore_id is a source compatibilty issue (because the > symbol > > > > > moved to > > > > > a new location), which is irrelevant to us. Its not in any way a > > > binary > > > > > compat > > > > > problem, which is what we care about. Sorry for the noise. > > > > > > > > > > I do still have a few concerns about some changed calling > conventions > > > with > > > > > a few > > > > > other functions, which I'll look into on monday. > > > > > > > > > > > > > > Please let me know your inputs on changed calling conventions. Most > of > > > them > > > > can be fixed by re-arranging moved code in _common_ files and order > of > > > > compilation. > > > > > > > If moving the order of compliation around fixes the problem, then I am > > > reasonably convinced that it is, if not a false positive, a minor issue > > > with the > > > compilers dwarf information (The compiler just can't sanely change the > > > location > > > in which parameters are passed). If you make those changes, I'll ACK > > > them, and > > > look into whats going on with the calling conventions > > > > > > > Issues like the one shown below are taken care by reordering the code > > compilation. > > > > eal_parse_sysfs_value ( char const* filename, unsigned long* val ) > > > > Change > > The parameter filename became passed on stack instead of rdi register > > > > Effect > > Violation of the calling convention. This may result in crash or > incorrect > > behavior of applications. > > > > Last one that is left out is in > > > > rte_thread_set_affinity ( rte_cpuset_t* p1 ) > > > > Change > > The parameter *p1* became passed in *rdi* register instead of stack. > > > > Effect > > Violation of the calling convention. This may result in crash or > incorrect > > behavior of applications. > > > > After checking abi-0.99.pdf (x86-64.org) looks like for > > "rte_thread_set_affinity" new code is doing the right thing by passing > the > > parameter in "rdi" register since pointer is classified as > "integer_class". > > Nothing needs to be fixed here. After you confirm that warning can be > > ignored I will work on sending new revision. > > > ACK then, send the new revision, this appears to be a false positive. > > Thanks for taking the time to confirm. > Thanks Neil. I have sent v8 which fixes ABI warnings. I have tested it with x86_64-native-linuxapp-gcc, x86_64-native-linuxapp-clang and x86_64-ivshmem-gcc targets. ABI results look fine to me. I tried to run validate-abi.sh on BSD but ran into errors. If there is a way to check against BSD please let me know. > > Best > Neil > > > Thanks, > > Ravi > > > > > > > Thanks! > > > Neil > > > > > > > Thanks, > > > > Ravi > > > > > > > > Regards > > > > > Neil > > > > > > > > > > > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-28 23:52 ` Ravi Kerur @ 2015-04-29 10:04 ` Neil Horman 2015-04-29 17:47 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-29 10:04 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Tue, Apr 28, 2015 at 04:52:37PM -0700, Ravi Kerur wrote: > On Tue, Apr 28, 2015 at 12:35 PM, Neil Horman <nhorman@tuxdriver.com> wrote: > > > On Mon, Apr 27, 2015 at 03:39:41PM -0700, Ravi Kerur wrote: > > > On Mon, Apr 27, 2015 at 6:44 AM, Neil Horman <nhorman@tuxdriver.com> > > wrote: > > > > > > > On Sat, Apr 25, 2015 at 05:09:01PM -0700, Ravi Kerur wrote: > > > > > On Sat, Apr 25, 2015 at 6:02 AM, Neil Horman <nhorman@tuxdriver.com> > > > > wrote: > > > > > > > > > > > On Sat, Apr 25, 2015 at 08:32:42AM -0400, Neil Horman wrote: > > > > > > > On Fri, Apr 24, 2015 at 06:45:06PM -0700, Ravi Kerur wrote: > > > > > > > > On Fri, Apr 24, 2015 at 2:24 PM, Ravi Kerur <rkerur@gmail.com> > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Fri, Apr 24, 2015 at 12:51 PM, Neil Horman < > > > > nhorman@tuxdriver.com > > > > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > >> On Fri, Apr 24, 2015 at 12:21:23PM -0700, Ravi Kerur wrote: > > > > > > > > >> > On Fri, Apr 24, 2015 at 11:53 AM, Neil Horman < > > > > > > nhorman@tuxdriver.com> > > > > > > > > >> wrote: > > > > > > > > >> > > > > > > > > > >> > > On Fri, Apr 24, 2015 at 09:45:24AM -0700, Ravi Kerur > > wrote: > > > > > > > > >> > > > On Fri, Apr 24, 2015 at 8:22 AM, Neil Horman < > > > > > > nhorman@tuxdriver.com > > > > > > > > >> > > > > > > > > > >> > > wrote: > > > > > > > > >> > > > > > > > > > > > >> > > > > On Fri, Apr 24, 2015 at 08:14:04AM -0700, Ravi Kerur > > > > wrote: > > > > > > > > >> > > > > > On Fri, Apr 24, 2015 at 6:51 AM, Neil Horman < > > > > > > > > >> nhorman@tuxdriver.com> > > > > > > > > >> > > > > wrote: > > > > > > > > >> > > > > > > > > > > > > > >> > > > > > > On Thu, Apr 23, 2015 at 02:35:31PM -0700, Ravi > > Kerur > > > > > > wrote: > > > > > > > > >> > > > > > > > Changes in v7 > > > > > > > > >> > > > > > > > Remove _setname_ pthread calls. > > > > > > > > >> > > > > > > > Use rte_gettid() API in RTE_LOG to print > > > > thread_id. > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > Changes in v6 > > > > > > > > >> > > > > > > > Remove RTE_EXEC_ENV_BSDAPP from > > > > eal_common_thread.c > > > > > > file. > > > > > > > > >> > > > > > > > Add pthread_setname_np/pthread_set_name_np for > > > > > > Linux/FreeBSD > > > > > > > > >> > > > > > > > respectively. Plan to use _getname_ in RTE_LOG > > > > when > > > > > > > > >> available. > > > > > > > > >> > > > > > > > Use existing rte_get_systid() in RTE_LOG to > > print > > > > > > thread_id. > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > Changes in v5 > > > > > > > > >> > > > > > > > Rebase to latest code. > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > Changes in v4 > > > > > > > > >> > > > > > > > None > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > Changes in v3 > > > > > > > > >> > > > > > > > Changed subject to be more explicit on file > > name > > > > > > inclusion. > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > Changes in v2 > > > > > > > > >> > > > > > > > None > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > Changes in v1 > > > > > > > > >> > > > > > > > eal_thread.c has minor differences between > > Linux > > > > and > > > > > > BSD, > > > > > > > > >> move > > > > > > > > >> > > > > > > > entire file into common directory. > > > > > > > > >> > > > > > > > Use RTE_EXEC_ENV_BSDAPP to differentiate on > > minor > > > > > > > > >> differences. > > > > > > > > >> > > > > > > > Rename eal_thread.c to eal_common_thread.c > > > > > > > > >> > > > > > > > Makefile changes to reflect file move and name > > > > change. > > > > > > > > >> > > > > > > > Fix checkpatch warnings. > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > > > > > > > >> > > > > > > > --- > > > > > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/Makefile | > > 2 > > > > +- > > > > > > > > >> > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c | > > 152 > > > > > > > > >> > > > > > > ------------------------------ > > > > > > > > >> > > > > > > > lib/librte_eal/common/eal_common_thread.c | > > 147 > > > > > > > > >> > > > > > > ++++++++++++++++++++++++++++- > > > > > > > > >> > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c | > > 152 > > > > > > > > >> > > > > > > +----------------------------- > > > > > > > > >> > > > > > > > 4 files changed, 148 insertions(+), 305 > > > > deletions(-) > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > diff --git > > a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > > >> > > > > > > > index 2357cfa..55971b9 100644 > > > > > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/Makefile > > > > > > > > >> > > > > > > > @@ -87,7 +87,7 @@ CFLAGS_eal_common_log.o := > > > > > > -D_GNU_SOURCE > > > > > > > > >> > > > > > > > # workaround for a gcc bug with noreturn > > > > attribute > > > > > > > > >> > > > > > > > # > > > > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603 > > > > > > > > >> > > > > > > > ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) > > > > > > > > >> > > > > > > > -CFLAGS_eal_thread.o += -Wno-return-type > > > > > > > > >> > > > > > > > +CFLAGS_eal_common_thread.o += > > -Wno-return-type > > > > > > > > >> > > > > > > > CFLAGS_eal_hpet.o += -Wno-return-type > > > > > > > > >> > > > > > > > endif > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > diff --git > > > > a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > > >> > > > > > > b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > > >> > > > > > > > index 9a03437..5714b8f 100644 > > > > > > > > >> > > > > > > > --- a/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > > >> > > > > > > > +++ b/lib/librte_eal/bsdapp/eal/eal_thread.c > > > > > > > > >> > > > > > > > @@ -35,163 +35,11 @@ > > > > > > > > >> > > > > > > > #include <stdio.h> > > > > > > > > >> > > > > > > > #include <stdlib.h> > > > > > > > > >> > > > > > > > #include <stdint.h> > > > > > > > > >> > > > > > > > -#include <unistd.h> > > > > > > > > >> > > > > > > > -#include <sched.h> > > > > > > > > >> > > > > > > > -#include <pthread_np.h> > > > > > > > > >> > > > > > > > -#include <sys/queue.h> > > > > > > > > >> > > > > > > > #include <sys/thr.h> > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > -#include <rte_debug.h> > > > > > > > > >> > > > > > > > -#include <rte_atomic.h> > > > > > > > > >> > > > > > > > -#include <rte_launch.h> > > > > > > > > >> > > > > > > > -#include <rte_log.h> > > > > > > > > >> > > > > > > > -#include <rte_memory.h> > > > > > > > > >> > > > > > > > -#include <rte_memzone.h> > > > > > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > > > > > >> > > > > > > > -#include <rte_eal.h> > > > > > > > > >> > > > > > > > -#include <rte_per_lcore.h> > > > > > > > > >> > > > > > > > -#include <rte_lcore.h> > > > > > > > > >> > > > > > > > - > > > > > > > > >> > > > > > > > #include "eal_private.h" > > > > > > > > >> > > > > > > > #include "eal_thread.h" > > > > > > > > >> > > > > > > > > > > > > > > > >> > > > > > > > -RTE_DEFINE_PER_LCORE(unsigned, _lcore_id) = > > > > > > LCORE_ID_ANY; > > > > > > > > >> > > > > > > NAK, these are exported symbols, you can't > > remove > > > > them > > > > > > without > > > > > > > > >> > > going > > > > > > > > >> > > > > > > through the > > > > > > > > >> > > > > > > deprecation process. > > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > They are not removed/deleted, they are moved from > > > > > > eal_thread.c > > > > > > > > >> to > > > > > > > > >> > > > > > eal_common_thread.c file since it is common to > > both > > > > Linux > > > > > > and > > > > > > > > >> BSD. > > > > > > > > >> > > > > > > > > > > > > > >> > > > > Then perhaps you forgot to export the symbol? Its > > > > showing > > > > > > up as > > > > > > > > >> > > removed > > > > > > > > >> > > > > on the > > > > > > > > >> > > > > ABI checker utility. > > > > > > > > >> > > > > > > > > > > > > >> > > > > Neil > > > > > > > > >> > > > > > > > > > > > > >> > > > > > > > > > > > >> > > > Can you please show me in the current code where it is > > > > being > > > > > > > > >> exported? I > > > > > > > > >> > > > have only moved definitions to _common_ files, not > > sure > > > > why it > > > > > > > > >> should be > > > > > > > > >> > > > exported now. I searched in the current code for > > > > > > > > >> RTE_DEFINE_PER_LCORE > > > > > > > > >> > > > > > > > > > > > >> > > > #home/rkerur/dpdk-tmp/dpdk# grep -ir > > RTE_DEFINE_PER_LCORE > > > > * > > > > > > > > >> > > > app/test/test_per_lcore.c:static > > > > > > RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > > >> test) = > > > > > > > > >> > > > 0x12345678; > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > > > > lib/librte_eal/linuxapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > > > > > >> > > > _cpuset); > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > > >> > > > _lcore_id) = LCORE_ID_ANY; > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(unsigned, > > > > > > > > >> > > > _socket_id) = (unsigned)SOCKET_ID_ANY; > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > > > > > lib/librte_eal/bsdapp/eal/eal_thread.c:RTE_DEFINE_PER_LCORE(rte_cpuset_t, > > > > > > > > >> > > > _cpuset); > > > > > > > > >> > > > lib/librte_eal/common/include/rte_per_lcore.h:#define > > > > > > > > >> > > > RTE_DEFINE_PER_LCORE(type, name) \ > > > > > > > > >> > > > lib/librte_eal/common/include/rte_eal.h: static > > > > > > > > >> > > > RTE_DEFINE_PER_LCORE(int, _thread_id) = -1; > > > > > > > > >> > > > > > > > > > lib/librte_eal/common/eal_common_errno.c:RTE_DEFINE_PER_LCORE(int, > > > > > > > > >> > > > _rte_errno); > > > > > > > > >> > > > lib/librte_eal/common/eal_common_errno.c: static > > > > > > > > >> > > > RTE_DEFINE_PER_LCORE(char[RETVAL_SZ], retval); > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > > > > >> > > > > > Thanks > > > > > > > > >> > > > > > Ravi > > > > > > > > >> > > > > > > > > > > > > > >> > > > > > Regards > > > > > > > > >> > > > > > > Neil > > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > > > > > > > > >> > > > > > > > > > > > > >> > > Its exported in the version map file: > > > > > > > > >> > > per_lcore__lcore_id; > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > Thanks Neil, I checked and both linux and bsd > > > > rte_eal_version.map > > > > > > have > > > > > > > > >> it. > > > > > > > > >> > I compared .map file between "changed code" and the > > original, > > > > > > they are > > > > > > > > >> same > > > > > > > > >> > for both linux and bsd. In fact you had ACK'd v4 version > > of > > > > this > > > > > > patch > > > > > > > > >> > series and no major changes after that. Please let me > > know if > > > > I > > > > > > missed > > > > > > > > >> > something. > > > > > > > > >> > > > > > > > > > >> I did, and I'm retracting that, because I didn't think to > > check > > > > the > > > > > > ABI > > > > > > > > >> compatibility on this. But I ran it throught the ABI > > checking > > > > > > script > > > > > > > > >> this and > > > > > > > > >> this error popped out. You should run it as well, its in > > the > > > > > > scripts > > > > > > > > >> directory. > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> I see in your first patch you removed it and re-added it in > > the > > > > > > common > > > > > > > > >> section. > > > > > > > > >> But something about how its building is causing it to not > > show > > > > up > > > > > > as an > > > > > > > > >> exported > > > > > > > > >> symbol, which is problematic, as other applications are > > going to > > > > > > want > > > > > > > > >> access to > > > > > > > > >> it. > > > > > > > > >> > > > > > > > > >> It also possible that the ABI checker is throwing a false > > > > positive, > > > > > > but > > > > > > > > >> either > > > > > > > > >> way, it needs to be looked into prior to moving forward with > > > > this. > > > > > > > > >> > > > > > > > > >> > > > > > > > > > I did following things. > > > > > > > > > > > > > > > > > > Put a tag (v2.0.0-before-common-eal) before EAL common > > functions > > > > > > changes > > > > > > > > > for commit (3c0c807038ad642f4be7deb9370293c39d12f029 net: > > remove > > > > > > unneeded > > > > > > > > > include) > > > > > > > > > > > > > > > > > > Put a tag (v2.0.0-common-eal) after EAL common functions > > changes > > > > for > > > > > > > > > commit (25737e5a7212630a7b5d8ca756860a062f403789 Move common > > > > > > functions in > > > > > > > > > eal_pci.c) > > > > > > > > > > > > > > > > > > Ran validate-abi against x86_64-native-linuxapp-gcc and > > > > > > > > > > > > > > > > > > v2.0.0-rc3 and v2.0.0-before-common-eal, html report for > > > > > > librte_eal.so > > > > > > > > > shows removed symbols for "per_lcore__cpuset" > > > > > > > > > > > > > > > > > > v2.0.0-rc3 and v2.0.0-common-eal, html report for > > librte_eal.so > > > > shows > > > > > > > > > removed symbols for "per_lcore__cpuset" > > > > > > > > > > > > > > > > > > Removed symbol is different from what you have reported and > > in my > > > > > > case I > > > > > > > > > see it even before my commit. If you are interested I can > > unicast > > > > > > you html > > > > > > > > > report file. Please let me know how to proceed. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > I did some experiment and found some interesting things. I > > will > > > > take > > > > > > eal.c > > > > > > > > as an example > > > > > > > > > > > > > > > > eal.c is split into eal_common_sysfs.c eal_common_mem_cfg.c > > > > > > > > eal_common_proc_type.c and eal_common_app_usage.c. In > > > > > > linuxapp/eal/Makefile > > > > > > > > if I compile new files right after eal.c as shown below > > > > > > > > > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) := eal.c > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_sysfs.c > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_mem_cfg.c > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += > > eal_common_proc_type.c > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += > > eal_common_app_usage.c > > > > > > > > ... > > > > > > > > > > > > > > > > validate-abi results matches baseline. Instead if i place new > > > > _common_ > > > > > > > > files in common area in linuxapp/eal/Makefile as shown below > > > > > > > > > > > > > > > > # from common dir > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_memzone.c > > > > > > > > SRCS-$(CONFIG_RTE_LIBRTE_EAL_LINUXAPP) += eal_common_log.c > > > > > > > > ... > > > > > > > > > > > > > > > > validate-abi reports problem in binary compatibility and source > > > > > > > > compatiblity > > > > > > > > > > > > > > > > eal_filesystem.h, librte_eal.so.1 > > > > > > > > [+] eal_parse_sysfs_value ( char const* filename, unsigned > > long* > > > > val ) > > > > > > > > @@ DPDK_2.0 (2) > > > > > > > > > > > > > > > > I believe files in common and linuxapp directory are compiled > > same > > > > way > > > > > > so > > > > > > > > not sure why placement in makefile makes difference. > > > > > > > > > > > > > > > > Could this be false-positive from validate-abi script?? > > > > > > > > > > > > > > > It could be, yes. Though I'm more inclined to think that > > perhaps in > > > > the > > > > > > new > > > > > > > version of the code we're not generating ithe same dwarf > > information > > > > out > > > > > > of it. > > > > > > > In fact for some reason, I've checked both the build before and > > after > > > > > > your > > > > > > > patch series, and the exported CFLAGS aren't getting passed to > > the > > > > build > > > > > > > properly, implying that we're not building all the code in the > > > > validator > > > > > > with > > > > > > > the -g flag, which the validator need to function properly. I'm > > > > looking > > > > > > into > > > > > > > that > > > > > > > Neil > > > > > > > > > > > > > > > > > > > > Found the problem, I was stupidly reading the report incorrectly. > > The > > > > > > problem > > > > > > regarding _lcore_id is a source compatibilty issue (because the > > symbol > > > > > > moved to > > > > > > a new location), which is irrelevant to us. Its not in any way a > > > > binary > > > > > > compat > > > > > > problem, which is what we care about. Sorry for the noise. > > > > > > > > > > > > I do still have a few concerns about some changed calling > > conventions > > > > with > > > > > > a few > > > > > > other functions, which I'll look into on monday. > > > > > > > > > > > > > > > > > Please let me know your inputs on changed calling conventions. Most > > of > > > > them > > > > > can be fixed by re-arranging moved code in _common_ files and order > > of > > > > > compilation. > > > > > > > > > If moving the order of compliation around fixes the problem, then I am > > > > reasonably convinced that it is, if not a false positive, a minor issue > > > > with the > > > > compilers dwarf information (The compiler just can't sanely change the > > > > location > > > > in which parameters are passed). If you make those changes, I'll ACK > > > > them, and > > > > look into whats going on with the calling conventions > > > > > > > > > > Issues like the one shown below are taken care by reordering the code > > > compilation. > > > > > > eal_parse_sysfs_value ( char const* filename, unsigned long* val ) > > > > > > Change > > > The parameter filename became passed on stack instead of rdi register > > > > > > Effect > > > Violation of the calling convention. This may result in crash or > > incorrect > > > behavior of applications. > > > > > > Last one that is left out is in > > > > > > rte_thread_set_affinity ( rte_cpuset_t* p1 ) > > > > > > Change > > > The parameter *p1* became passed in *rdi* register instead of stack. > > > > > > Effect > > > Violation of the calling convention. This may result in crash or > > incorrect > > > behavior of applications. > > > > > > After checking abi-0.99.pdf (x86-64.org) looks like for > > > "rte_thread_set_affinity" new code is doing the right thing by passing > > the > > > parameter in "rdi" register since pointer is classified as > > "integer_class". > > > Nothing needs to be fixed here. After you confirm that warning can be > > > ignored I will work on sending new revision. > > > > > ACK then, send the new revision, this appears to be a false positive. > > > > Thanks for taking the time to confirm. > > > > Thanks Neil. I have sent v8 which fixes ABI warnings. I have tested it with > x86_64-native-linuxapp-gcc, x86_64-native-linuxapp-clang and > x86_64-ivshmem-gcc targets. ABI results look fine to me. > > I tried to run validate-abi.sh on BSD but ran into errors. If there is a > way to check against BSD please let me know. > The ABI checker should work on BSD as far as I know, since it only relies on dwarf information in the output binary. What errors are you seeing? Neil ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-29 10:04 ` Neil Horman @ 2015-04-29 17:47 ` Ravi Kerur 2015-04-30 16:00 ` Neil Horman 0 siblings, 1 reply; 27+ messages in thread From: Ravi Kerur @ 2015-04-29 17:47 UTC (permalink / raw) To: Neil Horman; +Cc: dev > > I tried to run validate-abi.sh on BSD but ran into errors. If there is a > > way to check against BSD please let me know. > > > The ABI checker should work on BSD as far as I know, since it only relies > on > dwarf information in the output binary. What errors are you seeing? > dpdk-bsd:/home/rkerur/dpdk-validate-abi-1/dpdk # sh ./scripts/validate-abi.sh v2.0.0-rc3 v2.0.0-abi x86_64-native-bsdapp-clang mktemp: illegal option -- p usage: mktemp [-d] [-q] [-t prefix] [-u] template ... mktemp [-d] [-q] [-u] -t prefix Cant find abi-compliance-checker utility abi-compliance-checker is installed as shown below. dpdk-bsd:/home/rkerur/dpdk-validate-abi-1/dpdk # pkg install devel/abi-compliance-checker Updating FreeBSD repository catalogue... FreeBSD repository is up-to-date. All repositories are up-to-date. Checking integrity... done (0 conflicting) The most recent version of packages are already installed > > Neil > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-29 17:47 ` Ravi Kerur @ 2015-04-30 16:00 ` Neil Horman 2015-05-01 0:15 ` Ravi Kerur 0 siblings, 1 reply; 27+ messages in thread From: Neil Horman @ 2015-04-30 16:00 UTC (permalink / raw) To: Ravi Kerur; +Cc: dev On Wed, Apr 29, 2015 at 10:47:04AM -0700, Ravi Kerur wrote: > > > I tried to run validate-abi.sh on BSD but ran into errors. If there is a > > > way to check against BSD please let me know. > > > > > The ABI checker should work on BSD as far as I know, since it only relies > > on > > dwarf information in the output binary. What errors are you seeing? > > > > dpdk-bsd:/home/rkerur/dpdk-validate-abi-1/dpdk # sh > ./scripts/validate-abi.sh v2.0.0-rc3 v2.0.0-abi x86_64-native-bsdapp-clang > mktemp: illegal option -- p Ah, bsd mktemp doesn't support the -p option. I'll see if I can fix that. > usage: mktemp [-d] [-q] [-t prefix] [-u] template ... > mktemp [-d] [-q] [-u] -t prefix > Cant find abi-compliance-checker utility > > abi-compliance-checker is installed as shown below. > > dpdk-bsd:/home/rkerur/dpdk-validate-abi-1/dpdk # pkg install > devel/abi-compliance-checker > Updating FreeBSD repository catalogue... > FreeBSD repository is up-to-date. > All repositories are up-to-date. > Checking integrity... done (0 conflicting) > The most recent version of packages are already installed > Whats the path for abi-compliance checker there? It would seem that the binary isn't in your path, as which isn't locating it. Neil > > > > > Neil > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
* Re: [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c 2015-04-30 16:00 ` Neil Horman @ 2015-05-01 0:15 ` Ravi Kerur 0 siblings, 0 replies; 27+ messages in thread From: Ravi Kerur @ 2015-05-01 0:15 UTC (permalink / raw) To: Neil Horman; +Cc: dev On Thu, Apr 30, 2015 at 9:00 AM, Neil Horman <nhorman@tuxdriver.com> wrote: > On Wed, Apr 29, 2015 at 10:47:04AM -0700, Ravi Kerur wrote: > > > > I tried to run validate-abi.sh on BSD but ran into errors. If there > is a > > > > way to check against BSD please let me know. > > > > > > > The ABI checker should work on BSD as far as I know, since it only > relies > > > on > > > dwarf information in the output binary. What errors are you seeing? > > > > > > > dpdk-bsd:/home/rkerur/dpdk-validate-abi-1/dpdk # sh > > ./scripts/validate-abi.sh v2.0.0-rc3 v2.0.0-abi > x86_64-native-bsdapp-clang > > mktemp: illegal option -- p > Ah, bsd mktemp doesn't support the -p option. I'll see if I can fix that. > I think there are couple of other issues I found freeBSD sed is different from Linux (GNU sed) and I get following errors with the script "sed 1 command c expects \ followed by text". I have to use gsed (GNU sed) in freeBSD to get rid of that error and similarly freeBSD uses gmake instead of make. I have made those minor changes and sending them with this email as an attachment. > > usage: mktemp [-d] [-q] [-t prefix] [-u] template ... > > mktemp [-d] [-q] [-u] -t prefix > > Cant find abi-compliance-checker utility > > > > abi-compliance-checker is installed as shown below. > > > > dpdk-bsd:/home/rkerur/dpdk-validate-abi-1/dpdk # pkg install > > devel/abi-compliance-checker > > Updating FreeBSD repository catalogue... > > FreeBSD repository is up-to-date. > > All repositories are up-to-date. > > Checking integrity... done (0 conflicting) > > The most recent version of packages are already installed > > > > Whats the path for abi-compliance checker there? It would seem that the > binary > isn't in your path, as which isn't locating it. > I am using regular freeBSD port install which doesn't install in any /usr/bin or /usr/local/bin. I finally decided to install both abi-dumper and abi-compliance-checker from source, compile and install it in correct directory. Above error is fixed after that, however, abi utilities use "eu-readelf" and I can't find that utility to install in freeBSD. I get following errors ERROR: can't find "eu-readelf" command freeBSD has only readelf. Please let me know if there is a way to get rid of this error. Thanks, Ravi > > > > > > > Neil > > > > > > > ^ permalink raw reply [flat|nested] 27+ messages in thread
end of thread, other threads:[~2015-05-01 0:15 UTC | newest] Thread overview: 27+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2015-04-23 21:35 [dpdk-dev] [PATCH v7 0/6] Move EAL common functions Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 2/6] Move common functions in eal.c Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 3/6] Move common functions in eal_lcore.c Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 4/6] Move common functions in eal_timer.c Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 5/6] Move common functions in eal_memory.c Ravi Kerur 2015-04-23 21:35 ` [dpdk-dev] [PATCH v7 6/6] Move common functions in eal_pci.c Ravi Kerur 2015-04-24 13:51 ` [dpdk-dev] [PATCH v7 1/6] Move common functions in eal_thread.c Neil Horman 2015-04-24 15:14 ` Ravi Kerur 2015-04-24 15:22 ` Neil Horman 2015-04-24 16:45 ` Ravi Kerur 2015-04-24 18:53 ` Neil Horman 2015-04-24 19:21 ` Ravi Kerur 2015-04-24 19:51 ` Neil Horman 2015-04-24 21:24 ` Ravi Kerur 2015-04-25 1:45 ` Ravi Kerur 2015-04-25 12:32 ` Neil Horman 2015-04-25 13:02 ` Neil Horman 2015-04-26 0:09 ` Ravi Kerur 2015-04-27 13:44 ` Neil Horman 2015-04-27 22:39 ` Ravi Kerur 2015-04-28 19:35 ` Neil Horman 2015-04-28 23:52 ` Ravi Kerur 2015-04-29 10:04 ` Neil Horman 2015-04-29 17:47 ` Ravi Kerur 2015-04-30 16:00 ` Neil Horman 2015-05-01 0:15 ` Ravi Kerur
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).