| Message ID | 20230202110058.130695-11-xuanzhuo@linux.alibaba.com (mailing list archive) |
|---|---|
| State | Changes Requested |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | virtio-net: support AF_XDP zero copy | expand |
Hi Xuan,
Thank you for the patch! Yet something to improve:
[auto build test ERROR on net-next/master]
[also build test ERROR on mst-vhost/linux-next linus/master v6.2-rc6 next-20230203]
[cannot apply to net/master]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Xuan-Zhuo/virtio_ring-virtqueue_add-support-premapped/20230202-190707
patch link: https://lore.kernel.org/r/20230202110058.130695-11-xuanzhuo%40linux.alibaba.com
patch subject: [PATCH 10/33] xsk: support virtio DMA map
config: i386-debian-10.3-kvm (https://download.01.org/0day-ci/archive/20230206/202302060542.IxBGSiKh-lkp@intel.com/config)
compiler: gcc-11 (Debian 11.3.0-8) 11.3.0
reproduce (this is a W=1 build):
# https://github.com/intel-lab-lkp/linux/commit/370aefebcea755f7c4c14e16f8dcb5540769fd26
git remote add linux-review https://github.com/intel-lab-lkp/linux
git fetch --no-tags linux-review Xuan-Zhuo/virtio_ring-virtqueue_add-support-premapped/20230202-190707
git checkout 370aefebcea755f7c4c14e16f8dcb5540769fd26
# save the config file
mkdir build_dir && cp config build_dir/.config
make W=1 O=build_dir ARCH=i386 olddefconfig
make W=1 O=build_dir ARCH=i386 SHELL=/bin/bash
If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@intel.com>
All errors (new ones prefixed by >>):
ld: net/xdp/xsk_buff_pool.o: in function `xp_alloc':
>> net/xdp/xsk_buff_pool.c:575: undefined reference to `is_virtio_device'
>> ld: net/xdp/xsk_buff_pool.c:576: undefined reference to `virtio_dma_sync_signle_range_for_device'
ld: net/xdp/xsk_buff_pool.o: in function `__xp_dma_unmap':
net/xdp/xsk_buff_pool.c:338: undefined reference to `is_virtio_device'
>> ld: net/xdp/xsk_buff_pool.c:339: undefined reference to `virtio_dma_unmap'
ld: net/xdp/xsk_buff_pool.o: in function `xp_dma_map':
net/xdp/xsk_buff_pool.c:443: undefined reference to `is_virtio_device'
ld: net/xdp/xsk_buff_pool.c:443: undefined reference to `virtio_dma_sync_signle_range_for_device'
>> ld: net/xdp/xsk_buff_pool.c:443: undefined reference to `virtio_dma_sync_signle_range_for_cpu'
>> ld: net/xdp/xsk_buff_pool.c:458: undefined reference to `virtio_dma_map_page'
>> ld: net/xdp/xsk_buff_pool.c:461: undefined reference to `virtio_dma_mapping_error'
>> ld: net/xdp/xsk_buff_pool.c:464: undefined reference to `virtio_dma_need_sync'
>> ld: net/xdp/xsk_buff_pool.c:457: undefined reference to `is_virtio_device'
vim +575 net/xdp/xsk_buff_pool.c
424
425 int xp_dma_map(struct xsk_buff_pool *pool, struct device *dev,
426 unsigned long attrs, struct page **pages, u32 nr_pages)
427 {
428 struct xsk_dma_map *dma_map;
429 dma_addr_t dma;
430 int err;
431 u32 i;
432
433 dma_map = xp_find_dma_map(pool);
434 if (dma_map) {
435 err = xp_init_dma_info(pool, dma_map);
436 if (err)
437 return err;
438
439 refcount_inc(&dma_map->users);
440 return 0;
441 }
442
> 443 if (is_virtio_device(dev)) {
444 pool->dma_sync_for_cpu = virtio_dma_sync_signle_range_for_cpu;
445 pool->dma_sync_for_device = virtio_dma_sync_signle_range_for_device;
446
447 } else {
448 pool->dma_sync_for_cpu = dma_sync_for_cpu;
449 pool->dma_sync_for_device = dma_sync_for_device;
450 }
451
452 dma_map = xp_create_dma_map(dev, pool->netdev, nr_pages, pool->umem);
453 if (!dma_map)
454 return -ENOMEM;
455
456 for (i = 0; i < dma_map->dma_pages_cnt; i++) {
> 457 if (is_virtio_device(dev)) {
> 458 dma = virtio_dma_map_page(dev, pages[i], 0, PAGE_SIZE,
459 DMA_BIDIRECTIONAL);
460
> 461 if (virtio_dma_mapping_error(dev, dma))
462 goto err;
463
> 464 if (virtio_dma_need_sync(dev, dma))
465 dma_map->dma_need_sync = true;
466
467 } else {
468 dma = dma_map_page_attrs(dev, pages[i], 0, PAGE_SIZE,
469 DMA_BIDIRECTIONAL, attrs);
470
471 if (dma_mapping_error(dev, dma))
472 goto err;
473
474 if (dma_need_sync(dev, dma))
475 dma_map->dma_need_sync = true;
476 }
477 dma_map->dma_pages[i] = dma;
478 }
479
480 if (pool->unaligned)
481 xp_check_dma_contiguity(dma_map);
482
483 err = xp_init_dma_info(pool, dma_map);
484 if (err) {
485 __xp_dma_unmap(dma_map, attrs);
486 return err;
487 }
488
489 return 0;
490 err:
491 __xp_dma_unmap(dma_map, attrs);
492 return -ENOMEM;
493 }
494 EXPORT_SYMBOL(xp_dma_map);
495
496 static bool xp_addr_crosses_non_contig_pg(struct xsk_buff_pool *pool,
497 u64 addr)
498 {
499 return xp_desc_crosses_non_contig_pg(pool, addr, pool->chunk_size);
500 }
501
502 static bool xp_check_unaligned(struct xsk_buff_pool *pool, u64 *addr)
503 {
504 *addr = xp_unaligned_extract_addr(*addr);
505 if (*addr >= pool->addrs_cnt ||
506 *addr + pool->chunk_size > pool->addrs_cnt ||
507 xp_addr_crosses_non_contig_pg(pool, *addr))
508 return false;
509 return true;
510 }
511
512 static bool xp_check_aligned(struct xsk_buff_pool *pool, u64 *addr)
513 {
514 *addr = xp_aligned_extract_addr(pool, *addr);
515 return *addr < pool->addrs_cnt;
516 }
517
518 static struct xdp_buff_xsk *__xp_alloc(struct xsk_buff_pool *pool)
519 {
520 struct xdp_buff_xsk *xskb;
521 u64 addr;
522 bool ok;
523
524 if (pool->free_heads_cnt == 0)
525 return NULL;
526
527 for (;;) {
528 if (!xskq_cons_peek_addr_unchecked(pool->fq, &addr)) {
529 pool->fq->queue_empty_descs++;
530 return NULL;
531 }
532
533 ok = pool->unaligned ? xp_check_unaligned(pool, &addr) :
534 xp_check_aligned(pool, &addr);
535 if (!ok) {
536 pool->fq->invalid_descs++;
537 xskq_cons_release(pool->fq);
538 continue;
539 }
540 break;
541 }
542
543 if (pool->unaligned) {
544 xskb = pool->free_heads[--pool->free_heads_cnt];
545 xp_init_xskb_addr(xskb, pool, addr);
546 if (pool->dma_pages_cnt)
547 xp_init_xskb_dma(xskb, pool, pool->dma_pages, addr);
548 } else {
549 xskb = &pool->heads[xp_aligned_extract_idx(pool, addr)];
550 }
551
552 xskq_cons_release(pool->fq);
553 return xskb;
554 }
555
556 struct xdp_buff *xp_alloc(struct xsk_buff_pool *pool)
557 {
558 struct xdp_buff_xsk *xskb;
559
560 if (!pool->free_list_cnt) {
561 xskb = __xp_alloc(pool);
562 if (!xskb)
563 return NULL;
564 } else {
565 pool->free_list_cnt--;
566 xskb = list_first_entry(&pool->free_list, struct xdp_buff_xsk,
567 free_list_node);
568 list_del_init(&xskb->free_list_node);
569 }
570
571 xskb->xdp.data = xskb->xdp.data_hard_start + XDP_PACKET_HEADROOM;
572 xskb->xdp.data_meta = xskb->xdp.data;
573
574 if (pool->dma_need_sync) {
> 575 if (is_virtio_device(pool->dev))
> 576 virtio_dma_sync_signle_range_for_device(pool->dev, xskb->dma, 0,
577 pool->frame_len,
578 DMA_BIDIRECTIONAL);
579 else
580 dma_sync_single_range_for_device(pool->dev, xskb->dma, 0,
581 pool->frame_len,
582 DMA_BIDIRECTIONAL);
583 }
584 return &xskb->xdp;
585 }
586 EXPORT_SYMBOL(xp_alloc);
587
diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index 78e325e195fa..e2785aca8396 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -3,6 +3,7 @@ #include <net/xsk_buff_pool.h> #include <net/xdp_sock.h> #include <net/xdp_sock_drv.h> +#include <linux/virtio.h> #include "xsk_queue.h" #include "xdp_umem.h" @@ -334,8 +335,12 @@ static void __xp_dma_unmap(struct xsk_dma_map *dma_map, unsigned long attrs) dma = &dma_map->dma_pages[i]; if (*dma) { *dma &= ~XSK_NEXT_PG_CONTIG_MASK; - dma_unmap_page_attrs(dma_map->dev, *dma, PAGE_SIZE, - DMA_BIDIRECTIONAL, attrs); + if (is_virtio_device(dma_map->dev)) + virtio_dma_unmap(dma_map->dev, *dma, PAGE_SIZE, + DMA_BIDIRECTIONAL); + else + dma_unmap_page_attrs(dma_map->dev, *dma, PAGE_SIZE, + DMA_BIDIRECTIONAL, attrs); *dma = 0; } } @@ -435,22 +440,40 @@ int xp_dma_map(struct xsk_buff_pool *pool, struct device *dev, return 0; } - pool->dma_sync_for_cpu = dma_sync_for_cpu; - pool->dma_sync_for_device = dma_sync_for_device; + if (is_virtio_device(dev)) { + pool->dma_sync_for_cpu = virtio_dma_sync_signle_range_for_cpu; + pool->dma_sync_for_device = virtio_dma_sync_signle_range_for_device; + + } else { + pool->dma_sync_for_cpu = dma_sync_for_cpu; + pool->dma_sync_for_device = dma_sync_for_device; + } dma_map = xp_create_dma_map(dev, pool->netdev, nr_pages, pool->umem); if (!dma_map) return -ENOMEM; for (i = 0; i < dma_map->dma_pages_cnt; i++) { - dma = dma_map_page_attrs(dev, pages[i], 0, PAGE_SIZE, - DMA_BIDIRECTIONAL, attrs); - if (dma_mapping_error(dev, dma)) { - __xp_dma_unmap(dma_map, attrs); - return -ENOMEM; + if (is_virtio_device(dev)) { + dma = virtio_dma_map_page(dev, pages[i], 0, PAGE_SIZE, + DMA_BIDIRECTIONAL); + + if (virtio_dma_mapping_error(dev, dma)) + goto err; + + if (virtio_dma_need_sync(dev, dma)) + dma_map->dma_need_sync = true; + + } else { + dma = dma_map_page_attrs(dev, pages[i], 0, PAGE_SIZE, + DMA_BIDIRECTIONAL, attrs); + + if (dma_mapping_error(dev, dma)) + goto err; + + if (dma_need_sync(dev, dma)) + dma_map->dma_need_sync = true; } - if (dma_need_sync(dev, dma)) - dma_map->dma_need_sync = true; dma_map->dma_pages[i] = dma; } @@ -464,6 +487,9 @@ int xp_dma_map(struct xsk_buff_pool *pool, struct device *dev, } return 0; +err: + __xp_dma_unmap(dma_map, attrs); + return -ENOMEM; } EXPORT_SYMBOL(xp_dma_map); @@ -546,9 +572,14 @@ struct xdp_buff *xp_alloc(struct xsk_buff_pool *pool) xskb->xdp.data_meta = xskb->xdp.data; if (pool->dma_need_sync) { - dma_sync_single_range_for_device(pool->dev, xskb->dma, 0, - pool->frame_len, - DMA_BIDIRECTIONAL); + if (is_virtio_device(pool->dev)) + virtio_dma_sync_signle_range_for_device(pool->dev, xskb->dma, 0, + pool->frame_len, + DMA_BIDIRECTIONAL); + else + dma_sync_single_range_for_device(pool->dev, xskb->dma, 0, + pool->frame_len, + DMA_BIDIRECTIONAL); } return &xskb->xdp; }
When device is a virtio device, use virtio's DMA interface. Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com> --- net/xdp/xsk_buff_pool.c | 59 +++++++++++++++++++++++++++++++---------- 1 file changed, 45 insertions(+), 14 deletions(-)