search for: __gfp_zero

...the hypervisor and shouldn't be given to the host ksmd to scan. Therefore, it is not necessary to zero ballooned pages, which is very time consuming when the page amount is large. The ongoing fast balloon tests show that the time to balloon 7G pages is increased from ~491ms to 2.8 seconds with __GFP_ZERO added. So, this patch removes the flag. Signed-off-by: Wei Wang <wei.w.wang at intel.com> Cc: Michal Hocko <mhocko at kernel.org> Cc: Michael S. Tsirkin <mst at redhat.com> --- mm/balloon_compaction.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/balloon...

...the hypervisor and shouldn't be given to the host ksmd to scan. Therefore, it is not necessary to zero ballooned pages, which is very time consuming when the page amount is large. The ongoing fast balloon tests show that the time to balloon 7G pages is increased from ~491ms to 2.8 seconds with __GFP_ZERO added. So, this patch removes the flag. Signed-off-by: Wei Wang <wei.w.wang at intel.com> Cc: Michal Hocko <mhocko at kernel.org> Cc: Michael S. Tsirkin <mst at redhat.com> --- mm/balloon_compaction.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/balloon...

...the hypervisor and shouldn't be given to the host ksmd to scan. Therefore, it is not necessary to zero ballooned pages, which is very time consuming when the page amount is large. The ongoing fast balloon tests show that the time to balloon 7G pages is increased from ~491ms to 2.8 seconds with __GFP_ZERO added. So, this patch removes the flag. Signed-off-by: Wei Wang <wei.w.wang at intel.com> --- mm/balloon_compaction.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/balloon_compaction.c b/mm/balloon_compaction.c index 9075aa5..b06d9fe 100644 --- a/mm/balloon_compacti...

...the hypervisor and shouldn't be given to the host ksmd to scan. Therefore, it is not necessary to zero ballooned pages, which is very time consuming when the page amount is large. The ongoing fast balloon tests show that the time to balloon 7G pages is increased from ~491ms to 2.8 seconds with __GFP_ZERO added. So, this patch removes the flag. Signed-off-by: Wei Wang <wei.w.wang at intel.com> --- mm/balloon_compaction.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/balloon_compaction.c b/mm/balloon_compaction.c index 9075aa5..b06d9fe 100644 --- a/mm/balloon_compacti...

...000000000 -0700 +++ linux-2.6.14-rc1/arch/i386/kernel/smpboot.c 2005-09-28 12:54:08.000000000 -0700 @@ -898,7 +898,8 @@ static int __devinit do_boot_cpu(int api * This grunge runs the startup process for * the targeted processor. */ - cpu_gdt_descr[cpu].address = __get_free_page(GFP_KERNEL|__GFP_ZERO); + if (!cpu_gdt_descr[cpu].address) + cpu_gdt_descr[cpu].address = __get_free_page(GFP_KERNEL|__GFP_ZERO); atomic_set(&init_deasserted, 0);

...000000000 -0700 +++ linux-2.6.14-rc1/arch/i386/kernel/smpboot.c 2005-09-28 12:54:08.000000000 -0700 @@ -898,7 +898,8 @@ static int __devinit do_boot_cpu(int api * This grunge runs the startup process for * the targeted processor. */ - cpu_gdt_descr[cpu].address = __get_free_page(GFP_KERNEL|__GFP_ZERO); + if (!cpu_gdt_descr[cpu].address) + cpu_gdt_descr[cpu].address = __get_free_page(GFP_KERNEL|__GFP_ZERO); atomic_set(&init_deasserted, 0);

We used to have use a single parent for all DMA operations. This tends to complicate the mdev based hardware virtio datapath offloading which may not implement the control path over datapath like ctrl vq in the case of virtio-net. So this series tries to intorduce per DMA domain by allowing trasnport to specify the parent device for each virtqueue. Then for the case of virtio-mdev device, it can

...us_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size, newchannel->channel_callback_context = context; /* Allocate the ring buffer */ - out = osd_page_alloc((send_ringbuffer_size + recv_ringbuffer_size) - >> PAGE_SHIFT); + out = (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO, + get_order(send_ringbuffer_size + recv_ringbuffer_size)); + if (!out) return -ENOMEM; @@ -300,8 +301,8 @@ Cleanup: errorout: ringbuffer_cleanup(&newchannel->outbound); ringbuffer_cleanup(&newchannel->inbound); - osd_page_free(out, (send_ringbuffer_size + recv_ringbuffer...

...us_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size, newchannel->channel_callback_context = context; /* Allocate the ring buffer */ - out = osd_page_alloc((send_ringbuffer_size + recv_ringbuffer_size) - >> PAGE_SHIFT); + out = (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO, + get_order(send_ringbuffer_size + recv_ringbuffer_size)); + if (!out) return -ENOMEM; @@ -300,8 +301,8 @@ Cleanup: errorout: ringbuffer_cleanup(&newchannel->outbound); ringbuffer_cleanup(&newchannel->inbound); - osd_page_free(out, (send_ringbuffer_size + recv_ringbuffer...

...st); if (exbuf->num_bo_handles) { - bo_handles = kvmalloc_array(exbuf->num_bo_handles, - sizeof(uint32_t), GFP_KERNEL); + sizeof(uint32_t), GFP_KERNEL); buflist = kvmalloc_array(exbuf->num_bo_handles, - sizeof(struct ttm_validate_buffer), - GFP_KERNEL | __GFP_ZERO); + sizeof(struct drm_gem_object*), + GFP_KERNEL | __GFP_ZERO); if (!bo_handles || !buflist) { ret = -ENOMEM; goto out_unused_fd; @@ -181,19 +179,15 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data, ret = -ENOENT; goto out_unused_fd; }...

...st); if (exbuf->num_bo_handles) { - bo_handles = kvmalloc_array(exbuf->num_bo_handles, - sizeof(uint32_t), GFP_KERNEL); + sizeof(uint32_t), GFP_KERNEL); buflist = kvmalloc_array(exbuf->num_bo_handles, - sizeof(struct ttm_validate_buffer), - GFP_KERNEL | __GFP_ZERO); + sizeof(struct drm_gem_object*), + GFP_KERNEL | __GFP_ZERO); if (!bo_handles || !buflist) { ret = -ENOMEM; goto out_unused_fd; @@ -181,19 +179,15 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data, ret = -ENOENT; goto out_unused_fd; }...

...;t be given to the host ksmd to scan. I find MADV_DONTNEED reference still quite confusing. What do you think about the following wording instead: " Zeroying ballon pages is rather time consuming, especially when a lot of pages are in flight. E.g. 7GB worth of ballooned memory takes 2.8s with __GFP_ZERO while it takes ~491ms without it. The original commit argued that zeroying will help ksmd to merge these pages on the host but this argument is assuming that the host actually marks balloon pages for ksm which is not universally true. So we pay performance penalty for something that even might not...

...e Perches <joe at perches.com> wrote: > > On Mon, 2020-11-02 at 16:20 +0100, Bartosz Golaszewski wrote: > > > From: Bartosz Golaszewski <bgolaszewski at baylibre.com> > Yeah so I had this concern for devm_krealloc() and even sent a patch > that extended it to honor __GFP_ZERO before I noticed that regular > krealloc() silently ignores __GFP_ZERO. I'm not sure if this is on > purpose. Maybe we should either make krealloc() honor __GFP_ZERO or > explicitly state in its documentation that it ignores it? And my voice here is to ignore for the same reasons: res...

...vers/gpu/drm/nouveau/nouveau_dmem.c @@ -193,7 +193,7 @@ static vm_fault_t nouveau_dmem_migrate_to_ram(struct vm_fault *vmf) if (!spage || !(src & MIGRATE_PFN_MIGRATE)) goto done; - dpage = alloc_page_vma(GFP_HIGHUSER, vmf->vma, vmf->address); + dpage = alloc_page_vma(GFP_HIGHUSER | __GFP_ZERO, vmf->vma, vmf->address); if (!dpage) goto done; -- 2.34.1

...> + if (!gfpflags_allow_blocking(gfp)) { > + struct page *page; > + > + page = dma_common_alloc_pages(dev, size, handle, dir, gfp); > + if (!page) > + return NULL; > + return page_address(page); > + } > + > + return iommu_dma_alloc_remap(dev, size, handle, gfp | __GFP_ZERO, > + PAGE_KERNEL, 0); iommu_dma_alloc_remap() makes use of the DMA_ATTR_ALLOC_SINGLE_PAGES attribute to optimize the allocations for devices which don't care about how contiguous the backing memory is. Do you think we could add an attrs argument to this function and pass it there?...

...(struct virtio_device *vdev, unsigned index, > > info->num = num; > info->msix_vector = msix_vec; > + info->use_dma_api = vp_use_dma_api(); > > - size = PAGE_ALIGN(vring_size(num, VIRTIO_PCI_VRING_ALIGN)); > - info->queue = alloc_pages_exact(size, GFP_KERNEL|__GFP_ZERO); > + size = vring_size(num, VIRTIO_PCI_VRING_ALIGN); > + if (info->use_dma_api) { > + info->queue = dma_zalloc_coherent(vdev->dev.parent, size, > + &info->queue_dma_addr, > + GFP_KERNEL); > + } else { > + info->queue = alloc_pages_exact(PAGE_...

...(struct virtio_device *vdev, unsigned index, > > info->num = num; > info->msix_vector = msix_vec; > + info->use_dma_api = vp_use_dma_api(); > > - size = PAGE_ALIGN(vring_size(num, VIRTIO_PCI_VRING_ALIGN)); > - info->queue = alloc_pages_exact(size, GFP_KERNEL|__GFP_ZERO); > + size = vring_size(num, VIRTIO_PCI_VRING_ALIGN); > + if (info->use_dma_api) { > + info->queue = dma_zalloc_coherent(vdev->dev.parent, size, > + &info->queue_dma_addr, > + GFP_KERNEL); > + } else { > + info->queue = alloc_pages_exact(PAGE_...

...(struct virtio_device *vdev, unsigned index, > > info->num = num; > info->msix_vector = msix_vec; > + info->use_dma_api = vp_use_dma_api(); > > - size = PAGE_ALIGN(vring_size(num, VIRTIO_PCI_VRING_ALIGN)); > - info->queue = alloc_pages_exact(size, GFP_KERNEL|__GFP_ZERO); > + size = vring_size(num, VIRTIO_PCI_VRING_ALIGN); > + if (info->use_dma_api) { > + info->queue = dma_zalloc_coherent(vdev->dev.parent, size, > + &info->queue_dma_addr, > + GFP_KERNEL); > + } else { > + info->queue = alloc_pages_exact(PAGE_...

...(struct virtio_device *vdev, unsigned index, > > info->num = num; > info->msix_vector = msix_vec; > + info->use_dma_api = vp_use_dma_api(); > > - size = PAGE_ALIGN(vring_size(num, VIRTIO_PCI_VRING_ALIGN)); > - info->queue = alloc_pages_exact(size, GFP_KERNEL|__GFP_ZERO); > + size = vring_size(num, VIRTIO_PCI_VRING_ALIGN); > + if (info->use_dma_api) { > + info->queue = dma_zalloc_coherent(vdev->dev.parent, size, > + &info->queue_dma_addr, > + GFP_KERNEL); > + } else { > + info->queue = alloc_pages_exact(PAGE_...

...where this MADV_DONTNEED is done, please? > Therefore, it is not > necessary to zero ballooned pages, which is very time consuming when > the page amount is large. The ongoing fast balloon tests show that the > time to balloon 7G pages is increased from ~491ms to 2.8 seconds with > __GFP_ZERO added. So, this patch removes the flag. Please make it obvious that this is a revert of bb01b64cfab7 ("mm/balloon_compaction.c: enqueue zero page to balloon device"). > Signed-off-by: Wei Wang <wei.w.wang at intel.com> > --- > mm/balloon_compaction.c | 2 +- > 1 file...