search for: page_address

On Wed, Jan 15, 2014 at 09:36:13AM +0100, Geert Uytterhoeven wrote: > On Wed, Jan 15, 2014 at 9:26 AM, Michael S. Tsirkin <mst at redhat.com> wrote: > > As a result of backporting a bugfix, virtio_net started passing void * > > to page_address, assuming that it will get silently converted to struct > > page *. But this does not happen on architectures where page_address is > > a macro, the result is build failure as the macro tries to dereference > > void*. > > > > Fix by reordering code slightly, so we alwa...

On Wed, Jan 15, 2014 at 09:36:13AM +0100, Geert Uytterhoeven wrote: > On Wed, Jan 15, 2014 at 9:26 AM, Michael S. Tsirkin <mst at redhat.com> wrote: > > As a result of backporting a bugfix, virtio_net started passing void * > > to page_address, assuming that it will get silently converted to struct > > page *. But this does not happen on architectures where page_address is > > a macro, the result is build failure as the macro tries to dereference > > void*. > > > > Fix by reordering code slightly, so we alwa...

As a result of backporting a bugfix, virtio_net started passing void * to page_address, assuming that it will get silently converted to struct page *. But this does not happen on architectures where page_address is a macro, the result is build failure as the macro tries to dereference void*. Fix by reordering code slightly, so we always pass struct page * to page_address. Reported-...

As a result of backporting a bugfix, virtio_net started passing void * to page_address, assuming that it will get silently converted to struct page *. But this does not happen on architectures where page_address is a macro, the result is build failure as the macro tries to dereference void*. Fix by reordering code slightly, so we always pass struct page * to page_address. Reported-...

...+ } else if (unlikely(headroom < virtnet_get_headroom(vi))) { + xdp_room = SKB_DATA_ALIGN(VIRTIO_XDP_HEADROOM + + sizeof(struct skb_shared_info)); + if (*len + xdp_room > PAGE_SIZE) + return NULL; + + xdp_page = alloc_page(GFP_ATOMIC); + if (!xdp_page) + return NULL; + + memcpy(page_address(xdp_page) + VIRTIO_XDP_HEADROOM, + page_address(*page) + offset, *len); + } else { + return page_address(*page) + offset; + } + + *frame_sz = PAGE_SIZE; + + put_page(*page); + + *page = xdp_page; + + return page_address(xdp_page) + VIRTIO_XDP_HEADROOM; +} + static struct sk_buff *receive_...

...adroom(vi))) { > + xdp_room = SKB_DATA_ALIGN(VIRTIO_XDP_HEADROOM + > + sizeof(struct skb_shared_info)); > + if (*len + xdp_room > PAGE_SIZE) > + return NULL; > + > + xdp_page = alloc_page(GFP_ATOMIC); > + if (!xdp_page) > + return NULL; > + > + memcpy(page_address(xdp_page) + VIRTIO_XDP_HEADROOM, > + page_address(*page) + offset, *len); > + } else { > + return page_address(*page) + offset; This makes the code a little harder to be read than the original code. Why not do a verbatim moving without introducing new logic? (Or introducing ne...

...ve_queue *rq, > + u16 *num_buf, > + struct page *p, > + int offset, > + int page_off, > + unsigned int *len) > +{ > + struct page *page = alloc_page(GFP_ATOMIC); > + > + if (!page) > + return NULL; > + > + memcpy(page_address(page) + page_off, page_address(p) + offset, *len); > + page_off += *len; > + > + while (--*num_buf) { > + unsigned int buflen; > + void *buf; > + int off; > + > + buf = virtqueue_get_buf(rq->vq, &buflen); > + if (unlikely(!buf)) > + goto err_buf; > + &...

...ve_queue *rq, > + u16 *num_buf, > + struct page *p, > + int offset, > + int page_off, > + unsigned int *len) > +{ > + struct page *page = alloc_page(GFP_ATOMIC); > + > + if (!page) > + return NULL; > + > + memcpy(page_address(page) + page_off, page_address(p) + offset, *len); > + page_off += *len; > + > + while (--*num_buf) { > + unsigned int buflen; > + void *buf; > + int off; > + > + buf = virtqueue_get_buf(rq->vq, &buflen); > + if (unlikely(!buf)) > + goto err_buf; > + &...

...g. + */ +static struct page *xdp_linearize_page(struct receive_queue *rq, + u16 *num_buf, + struct page *p, + int offset, + int page_off, + unsigned int *len) +{ + struct page *page = alloc_page(GFP_ATOMIC); + + if (!page) + return NULL; + + memcpy(page_address(page) + page_off, page_address(p) + offset, *len); + page_off += *len; + + while (--*num_buf) { + unsigned int buflen; + void *buf; + int off; + + buf = virtqueue_get_buf(rq->vq, &buflen); + if (unlikely(!buf)) + goto err_buf; + + p = virt_to_head_page(buf); + off = buf - page_addre...

...>io_addr + bank->size; transfered = 0; - bio_for_each_segment(vec, bio, idx) { - if (unlikely(phys_mem + vec->bv_len > phys_end)) { + bio_for_each_segment(vec, bio, iter) { + if (unlikely(phys_mem + vec.bv_len > phys_end)) { bio_io_error(bio); return; } - user_mem = page_address(vec->bv_page) + vec->bv_offset; + user_mem = page_address(vec.bv_page) + vec.bv_offset; if (bio_data_dir(bio) == READ) - memcpy(user_mem, (void *) phys_mem, vec->bv_len); + memcpy(user_mem, (void *) phys_mem, vec.bv_len); else - memcpy((void *) phys_mem, user_mem, vec->bv_l...

...>io_addr + bank->size; transfered = 0; - bio_for_each_segment(vec, bio, idx) { - if (unlikely(phys_mem + vec->bv_len > phys_end)) { + bio_for_each_segment(vec, bio, iter) { + if (unlikely(phys_mem + vec.bv_len > phys_end)) { bio_io_error(bio); return; } - user_mem = page_address(vec->bv_page) + vec->bv_offset; + user_mem = page_address(vec.bv_page) + vec.bv_offset; if (bio_data_dir(bio) == READ) - memcpy(user_mem, (void *) phys_mem, vec->bv_len); + memcpy(user_mem, (void *) phys_mem, vec.bv_len); else - memcpy((void *) phys_mem, user_mem, vec->bv_l...

...>io_addr + bank->size; transfered = 0; - bio_for_each_segment(vec, bio, idx) { - if (unlikely(phys_mem + vec->bv_len > phys_end)) { + bio_for_each_segment(vec, bio, iter) { + if (unlikely(phys_mem + vec.bv_len > phys_end)) { bio_io_error(bio); return; } - user_mem = page_address(vec->bv_page) + vec->bv_offset; + user_mem = page_address(vec.bv_page) + vec.bv_offset; if (bio_data_dir(bio) == READ) - memcpy(user_mem, (void *) phys_mem, vec->bv_len); + memcpy(user_mem, (void *) phys_mem, vec.bv_len); else - memcpy((void *) phys_mem, user_mem, vec->bv_l...

...m. - */ -static struct page *xdp_linearize_page(struct receive_queue *rq, - u16 *num_buf, - struct page *p, - int offset, - int page_off, - unsigned int *len) -{ - struct page *page = alloc_page(GFP_ATOMIC); - - if (!page) - return NULL; - - memcpy(page_address(page) + page_off, page_address(p) + offset, *len); - page_off += *len; - - while (--*num_buf) { - unsigned int buflen; - void *buf; - int off; - - buf = virtqueue_get_buf(rq->vq, &buflen); - if (unlikely(!buf)) - goto err_buf; - - p = virt_to_head_page(buf); - off = buf - page_addre...

...*/ static struct page *xdp_linearize_page(struct receive_queue *rq, - u16 num_buf, + u16 *num_buf, struct page *p, int offset, unsigned int *len) @@ -497,7 +497,7 @@ static struct page *xdp_linearize_page(struct receive_queue *rq, memcpy(page_address(page) + page_off, page_address(p) + offset, *len); page_off += *len; - while (--num_buf) { + while (--*num_buf) { unsigned int buflen; unsigned long ctx; void *buf; @@ -507,19 +507,22 @@ static struct page *xdp_linearize_page(struct receive_queue *rq, if (unlikely(!ctx)) goto er...

...7 Jul 25 02:15:02 vega2008 kernel: [<c046dfcf>] ? kernel_read+0x32/0x43 Jul 25 02:15:02 vega2008 kernel: [<c0490e60>] ? load_elf_binary+0x359/0x1152 Jul 25 02:15:02 vega2008 kernel: [<c045a6ee>] ? get_user_pages+0x2d5/0x35c Jul 25 02:15:02 vega2008 kernel: [<c04570d2>] ? page_address+0x78/0x98 Jul 25 02:15:02 vega2008 kernel: [<c045735a>] ? kmap_high+0x19/0x16b Jul 25 02:15:02 vega2008 kernel: [<c04570d2>] ? page_address+0x78/0x98 Jul 25 02:15:02 vega2008 kernel: [<c046dced>] ? copy_strings+0x169/0x173 Jul 25 02:15:02 vega2008 kernel: [<c046ddad>] ?...

Hi: This series tries to re-enable XDP_REDIRECT for mergeable buffer which was removed since commit 7324f5399b06 ("virtio_net: disable XDP_REDIRECT in receive_mergeable() case"). Main concerns are: - not enough tailroom was reserved which breaks cpumap - complex logic like EWMA and linearizing during XDP processing Fix those by: - reserve enough tailroom during refill - disable EWMA

Hi: This series tries to re-enable XDP_REDIRECT for mergeable buffer which was removed since commit 7324f5399b06 ("virtio_net: disable XDP_REDIRECT in receive_mergeable() case"). Main concerns are: - not enough tailroom was reserved which breaks cpumap - complex logic like EWMA and linearizing during XDP processing Fix those by: - reserve enough tailroom during refill - disable EWMA

...t;mst at redhat.com> wrote: > On Wed, Jan 15, 2014 at 09:36:13AM +0100, Geert Uytterhoeven wrote: >> On Wed, Jan 15, 2014 at 9:26 AM, Michael S. Tsirkin <mst at redhat.com> wrote: >> > As a result of backporting a bugfix, virtio_net started passing void * >> > to page_address, assuming that it will get silently converted to struct >> > page *. But this does not happen on architectures where page_address is >> > a macro, the result is build failure as the macro tries to dereference >> > void*. >> > >> > Fix by reordering code...

Hi: This series brings two optimizations for virtio-net XDP: - avoid reset during XDP set - turn off offloads on demand Please review. Thanks Jason Wang (5): virtio_ring: allow to store zero as the ctx virtio-net: pack headroom into ctx for mergeable buffer virtio-net: switch to use new ctx API for small buffer virtio-net: do not reset during XDP set virtio-net: switch off offloads

Hi: This series brings two optimizations for virtio-net XDP: - avoid reset during XDP set - turn off offloads on demand Please review. Thanks Jason Wang (5): virtio_ring: allow to store zero as the ctx virtio-net: pack headroom into ctx for mergeable buffer virtio-net: switch to use new ctx API for small buffer virtio-net: do not reset during XDP set virtio-net: switch off offloads