* [dpdk-dev] [PATCH]xen:support Dom0 driver for Linux kernel 3.13.0 and later
@ 2014-06-03 12:59 Jijiang Liu
2014-06-09 15:55 ` Thomas Monjalon
0 siblings, 1 reply; 2+ messages in thread
From: Jijiang Liu @ 2014-06-03 12:59 UTC (permalink / raw)
To: dev
Since Linux kernel version 3.13.0, the xen_create/destroy_contiguous_region() API has been changed,
and the first parameter is physical address in the API.
Signed-off-by: Jijiang Liu <jijiang.liu@intel.com>
Acked-by: Huawei Xie <huawei.xie@intel.com>
Tested-by: Heng Ding <hengx.ding@intel.com>
---
lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c | 24 ++++++++++++++++++++++-
1 files changed, 23 insertions(+), 1 deletions(-)
diff --git a/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c b/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
index c254300..a91c7ec 100644
--- a/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
+++ b/lib/librte_eal/linuxapp/xen_dom0/dom0_mm_misc.c
@@ -64,6 +64,7 @@
#include <linux/errno.h>
#include <linux/vmalloc.h>
#include <linux/mm.h>
+#include <linux/version.h>
#include <xen/xen.h>
#include <xen/page.h>
@@ -262,6 +263,7 @@ dom0_memory_free(uint32_t rsv_size)
for (i = 0; i < dom0_dev.num_bigblock * 2; i += 2) {
vstart = rsv_mm_info[i].vir_addr;
if (vstart) {
+ #if LINUX_VERSION_CODE < KERNEL_VERSION(3, 13, 0)
if (rsv_mm_info[i].exchange_flag)
xen_destroy_contiguous_region(vstart,
DOM0_CONTIG_NUM_ORDER);
@@ -269,6 +271,17 @@ dom0_memory_free(uint32_t rsv_size)
xen_destroy_contiguous_region(vstart +
DOM0_MEMBLOCK_SIZE,
DOM0_CONTIG_NUM_ORDER);
+ #else
+ if (rsv_mm_info[i].exchange_flag)
+ xen_destroy_contiguous_region(rsv_mm_info[i].pfn
+ * PAGE_SIZE,
+ DOM0_CONTIG_NUM_ORDER);
+ if (rsv_mm_info[i + 1].exchange_flag)
+ xen_destroy_contiguous_region(rsv_mm_info[i].pfn
+ * PAGE_SIZE + DOM0_MEMBLOCK_SIZE,
+ DOM0_CONTIG_NUM_ORDER);
+ #endif
+
size = DOM0_MEMBLOCK_SIZE * 2;
vaddr = vstart;
while (size > 0) {
@@ -381,6 +394,10 @@ dom0_memory_reserve(uint32_t rsv_size)
uint64_t pfn, vstart, vaddr;
uint32_t i, num_block, size, allocated_size = 0;
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(3, 13, 0)
+ dma_addr_t dma_handle;
+#endif
+
/* 2M as memory block */
num_block = rsv_size / SIZE_PER_BLOCK;
@@ -452,8 +469,13 @@ dom0_memory_reserve(uint32_t rsv_size)
* This API is used to exchage MFN for getting a block of
* contiguous physical addresses, its maximum size is 2M.
*/
+ #if LINUX_VERSION_CODE < KERNEL_VERSION(3, 13, 0)
if (xen_create_contiguous_region(rsv_mm_info[i].vir_addr,
- DOM0_CONTIG_NUM_ORDER, 0) == 0) {
+ DOM0_CONTIG_NUM_ORDER, 0) == 0) {
+ #else
+ if (xen_create_contiguous_region(rsv_mm_info[i].pfn * PAGE_SIZE,
+ DOM0_CONTIG_NUM_ORDER, 0, &dma_handle) == 0) {
+ #endif
rsv_mm_info[i].exchange_flag = 1;
rsv_mm_info[i].mfn =
pfn_to_mfn(rsv_mm_info[i].pfn);
--
1.7.7.6
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [dpdk-dev] [PATCH]xen:support Dom0 driver for Linux kernel 3.13.0 and later
2014-06-03 12:59 [dpdk-dev] [PATCH]xen:support Dom0 driver for Linux kernel 3.13.0 and later Jijiang Liu
@ 2014-06-09 15:55 ` Thomas Monjalon
0 siblings, 0 replies; 2+ messages in thread
From: Thomas Monjalon @ 2014-06-09 15:55 UTC (permalink / raw)
To: Jijiang Liu; +Cc: dev
2014-06-03 20:59, Jijiang Liu:
> Since Linux kernel version 3.13.0, the
> xen_create/destroy_contiguous_region() API has been changed, and the first
> parameter is physical address in the API.
>
> Signed-off-by: Jijiang Liu <jijiang.liu@intel.com>
> Acked-by: Huawei Xie <huawei.xie@intel.com>
> Tested-by: Heng Ding <hengx.ding@intel.com>
Applied for version 1.7.0.
Thanks
--
Thomas
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2014-06-09 15:55 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-06-03 12:59 [dpdk-dev] [PATCH]xen:support Dom0 driver for Linux kernel 3.13.0 and later Jijiang Liu
2014-06-09 15:55 ` Thomas Monjalon
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).