Re: [dpdk-dev] [RFC PATCH 1/7] eal: add linear thread id as pthread-local variable

["Qiu, Michael" <michael.qiu@intel.com>]

On 12/11/2014 10:05 AM, Cunming Liang wrote:
> Signed-off-by: Cunming Liang <cunming.liang@intel.com>
> ---
>  lib/librte_eal/common/include/rte_eal.h   |   5 ++
>  lib/librte_eal/common/include/rte_lcore.h |  12 ++++
>  lib/librte_eal/linuxapp/eal/eal_thread.c  | 115 ++++++++++++++++++++++++++++--
>  3 files changed, 126 insertions(+), 6 deletions(-)
>
> diff --git a/lib/librte_eal/common/include/rte_eal.h b/lib/librte_eal/common/include/rte_eal.h
> index f4ecd2e..2640167 100644
> --- a/lib/librte_eal/common/include/rte_eal.h
> +++ b/lib/librte_eal/common/include/rte_eal.h
> @@ -262,6 +262,11 @@ rte_set_application_usage_hook( rte_usage_hook_t usage_func );
>   */
>  int rte_eal_has_hugepages(void);
>  
> +#ifndef RTE_MAX_THREAD
> +#define RTE_MAX_THREAD                RTE_MAX_LCORE
> +#endif
> +
> +
>  #ifdef __cplusplus
>  }
>  #endif
> diff --git a/lib/librte_eal/common/include/rte_lcore.h b/lib/librte_eal/common/include/rte_lcore.h
> index 49b2c03..cd83d47 100644
> --- a/lib/librte_eal/common/include/rte_lcore.h
> +++ b/lib/librte_eal/common/include/rte_lcore.h
> @@ -73,6 +73,7 @@ struct lcore_config {
>  extern struct lcore_config lcore_config[RTE_MAX_LCORE];
>  
>  RTE_DECLARE_PER_LCORE(unsigned, _lcore_id); /**< Per core "core id". */
> +RTE_DECLARE_PER_LCORE(unsigned, _thread_id); /**< Per thread "linear tid". */
>  
>  /**
>   * Return the ID of the execution unit we are running on.
> @@ -86,6 +87,17 @@ rte_lcore_id(void)
>  }
>  
>  /**
> + * Return the linear thread ID of the cache unit we are running on.
> + * @return
> + *   core ID
> + */
> +static inline unsigned long
> +rte_linear_thread_id(void)
> +{
> +	return RTE_PER_LCORE(_thread_id);
> +}
> +
> +/**
>   * Get the id of the master lcore
>   *
>   * @return
> diff --git a/lib/librte_eal/linuxapp/eal/eal_thread.c b/lib/librte_eal/linuxapp/eal/eal_thread.c
> index 80a985f..52478d6 100644
> --- a/lib/librte_eal/linuxapp/eal/eal_thread.c
> +++ b/lib/librte_eal/linuxapp/eal/eal_thread.c
> @@ -39,6 +39,7 @@
>  #include <pthread.h>
>  #include <sched.h>
>  #include <sys/queue.h>
> +#include <string.h>
>  
>  #include <rte_debug.h>
>  #include <rte_atomic.h>
> @@ -51,12 +52,19 @@
>  #include <rte_eal.h>
>  #include <rte_per_lcore.h>
>  #include <rte_lcore.h>
> +#include <rte_spinlock.h>
> +#include <rte_common.h>
>  
>  #include "eal_private.h"
>  #include "eal_thread.h"
>  
> +#define LINEAR_THREAD_ID_POOL        "THREAD_ID_POOL"
> +
>  RTE_DEFINE_PER_LCORE(unsigned, _lcore_id);
>  
> +/* define linear thread id as thread-local variables */
> +RTE_DEFINE_PER_LCORE(unsigned, _thread_id);
> +
>  /*
>   * 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
> @@ -94,12 +102,13 @@ rte_eal_remote_launch(int (*f)(void *), void *arg, unsigned slave_id)
>  	return 0;
>  }
>  
> +
>  /* set affinity for current thread */
>  static int
> -eal_thread_set_affinity(void)
> +__eal_thread_set_affinity(pthread_t thread, unsigned lcore)
>  {
> +
>  	int s;
> -	pthread_t thread;
>  
>  /*
>   * According to the section VERSIONS of the CPU_ALLOC man page:
> @@ -126,9 +135,8 @@ eal_thread_set_affinity(void)
>  
>  	size = CPU_ALLOC_SIZE(RTE_MAX_LCORE);
>  	CPU_ZERO_S(size, cpusetp);
> -	CPU_SET_S(rte_lcore_id(), size, cpusetp);
> +	CPU_SET_S(lcore, size, cpusetp);
>  
> -	thread = pthread_self();
>  	s = pthread_setaffinity_np(thread, size, cpusetp);
>  	if (s != 0) {
>  		RTE_LOG(ERR, EAL, "pthread_setaffinity_np failed\n");
> @@ -140,9 +148,8 @@ eal_thread_set_affinity(void)
>  #else /* CPU_ALLOC */
>  	cpu_set_t cpuset;
>  	CPU_ZERO( &cpuset );
> -	CPU_SET( rte_lcore_id(), &cpuset );
> +	CPU_SET(lcore, &cpuset );
>  
> -	thread = pthread_self();
>  	s = pthread_setaffinity_np(thread, sizeof( cpuset ), &cpuset);
>  	if (s != 0) {
>  		RTE_LOG(ERR, EAL, "pthread_setaffinity_np failed\n");
> @@ -152,6 +159,15 @@ eal_thread_set_affinity(void)
>  	return 0;
>  }
>  
> +/* set affinity for current thread */
> +static int
> +eal_thread_set_affinity(void)
> +{
> +	pthread_t thread = pthread_self();
> +
> +	return __eal_thread_set_affinity(thread, rte_lcore_id());
> +}
> +
>  void eal_thread_init_master(unsigned lcore_id)
>  {
>  	/* set the lcore ID in per-lcore memory area */
> @@ -162,6 +178,87 @@ void eal_thread_init_master(unsigned lcore_id)
>  		rte_panic("cannot set affinity\n");
>  }
>  
> +/* linear thread id control block */
> +struct eal_thread_cb {
> +	rte_spinlock_t lock;
> +	uint64_t nb_bucket;
> +	uint64_t bitmap[0];
> +};
> +

Can this struct been declared in header files?
 
> +static struct eal_thread_cb *
> +__create_tid_pool(void)
> +{
> +	const struct rte_memzone *mz;
> +	struct eal_thread_cb *pcb;
> +	uint64_t sz;
> +	uint64_t nb_bucket;
> +
> +	nb_bucket = RTE_ALIGN_CEIL(RTE_MAX_THREAD, 64) / 64;

 Is it better to replace division to right shift?

nb_bucket = RTE_ALIGN_CEIL(RTE_MAX_THREAD, 64) >> 6;


> +	sz = sizeof(*pcb) + nb_bucket * sizeof(uint64_t);
> +	mz = rte_memzone_reserve(LINEAR_THREAD_ID_POOL,
> +				 sz, rte_socket_id(), 0);
> +	if (mz == NULL)
> +		rte_panic("Cannot allocate linear thread ID pool\n");
> +
> +	pcb = mz->addr;
> +	rte_spinlock_init(&pcb->lock);
> +	pcb->nb_bucket = nb_bucket;
> +	memset(pcb->bitmap, 0, nb_bucket * sizeof(uint64_t));
> +
> +	return pcb;
> +}
> +
> +static int
> +__get_linear_tid(uint64_t *tid)
> +{
> +	const struct rte_memzone *mz;
> +	struct eal_thread_cb *pcb;
> +	uint64_t i;
> +	uint8_t shift = 0;
> +
> +	mz = rte_memzone_lookup(LINEAR_THREAD_ID_POOL);
> +	if (mz != NULL)
> +		pcb = mz->addr;
> +	else
> +		pcb = __create_tid_pool();
> +
> +	rte_spinlock_lock(&pcb->lock);
> +	for (i = 0; i < pcb->nb_bucket; i++) {
> +		if (pcb->bitmap[i] == (uint64_t)-1)

It is better for bitmap as ~0(or ~(uint64_t)0) instead of (uint64_t)-1
for all bit set.
 
> +			continue;
> +		shift = 0; 
> +		while (pcb->bitmap[i] & (1UL << shift))
> +			shift ++;
> +		pcb->bitmap[i] |= (1UL << shift);
> +		break;
> +	}
> +	rte_spinlock_unlock(&pcb->lock);
> +
> +	if (i == pcb->nb_bucket)
> +		return -1;
> +
> +	*tid = i * 64 + shift;
> +	return 0;
> +}
> +
> +static void __rte_unused
> +__put_linear_tid(uint64_t tid)
> +{
> +	const struct rte_memzone *mz;
> +	struct eal_thread_cb *pcb;
> +	uint8_t shift;
> +
> +	mz = rte_memzone_lookup(LINEAR_THREAD_ID_POOL);
> +	if (!mz)
> +		return;
> +
> +	pcb = mz->addr;
> +	rte_spinlock_lock(&pcb->lock);
> +	shift = tid & 0x3F;
> +	pcb->bitmap[tid / 64] &= ~(1UL << shift);

tid >> 6


> +	rte_spinlock_unlock(&pcb->lock);	
> +}
> +
>  /* main loop of threads */
>  __attribute__((noreturn)) void *
>  eal_thread_loop(__attribute__((unused)) void *arg)
> @@ -169,6 +266,7 @@ eal_thread_loop(__attribute__((unused)) void *arg)
>  	char c;
>  	int n, ret;
>  	unsigned lcore_id;
> +	unsigned long ltid = 0;
>  	pthread_t thread_id;
>  	int m2s, s2m;
>  
> @@ -191,6 +289,11 @@ eal_thread_loop(__attribute__((unused)) void *arg)
>  	/* set the lcore ID in per-lcore memory area */
>  	RTE_PER_LCORE(_lcore_id) = lcore_id;
>  
> +	/* set the linear thread ID in per-lcore memory area */
> +	if (__get_linear_tid(&ltid) < 0)
> +		rte_panic("cannot get cache slot id\n");
> +	RTE_PER_LCORE(_thread_id) = ltid;
> +
>  	/* set CPU affinity */
>  	if (eal_thread_set_affinity() < 0)
>  		rte_panic("cannot set affinity\n");