All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>
To: dri-devel@lists.freedesktop.org
Cc: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>,
	"Andrew Morton" <akpm@linux-foundation.org>,
	"Matthew Wilcox (Oracle)" <willy@infradead.org>,
	"Miaohe Lin" <linmiaohe@huawei.com>,
	"David Hildenbrand" <david@redhat.com>,
	"Johannes Weiner" <hannes@cmpxchg.org>,
	"Peter Xu" <peterx@redhat.com>, NeilBrown <neilb@suse.de>,
	"Daniel Vetter" <daniel.vetter@ffwll.ch>,
	"Christian Koenig" <christian.koenig@amd.com>,
	"Dave Airlie" <airlied@redhat.com>,
	"Dave Hansen" <dave.hansen@intel.com>,
	"Matthew Auld" <matthew.auld@intel.com>,
	linux-graphics-maintainer@vmware.com, linux-mm@kvack.org,
	intel-gfx@lists.freedesktop.org
Subject: [RFC PATCH 13/16] drm/ttm: Make the call to ttm_tt_populate() interruptible when faulting
Date: Wed, 15 Feb 2023 17:14:02 +0100	[thread overview]
Message-ID: <20230215161405.187368-14-thomas.hellstrom@linux.intel.com> (raw)
In-Reply-To: <20230215161405.187368-1-thomas.hellstrom@linux.intel.com>

When swapping in, or under memory pressure ttm_tt_populate() may sleep
for a substantiable amount of time. Allow interrupts during the sleep.
This will also allow us to inject -EINTR errors during swapin in upcoming
patches.

Also avoid returning VM_FAULT_OOM, since that will confuse the core
mm, making it print out a confused message and retrying the fault.
Return VM_FAULT_SIGBUS also under OOM conditions.

Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
---
 drivers/gpu/drm/ttm/ttm_bo_vm.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 3ecda6db24b8..80f106bfe385 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -218,14 +218,21 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
 	prot = ttm_io_prot(bo, bo->resource, prot);
 	if (!bo->resource->bus.is_iomem) {
 		struct ttm_operation_ctx ctx = {
-			.interruptible = false,
+			.interruptible = true,
 			.no_wait_gpu = false,
 			.force_alloc = true
 		};
 
 		ttm = bo->ttm;
-		if (ttm_tt_populate(bdev, bo->ttm, &ctx))
-			return VM_FAULT_OOM;
+		err = ttm_tt_populate(bdev, bo->ttm, &ctx);
+		if (err) {
+			if (err == -EINTR || err == -ERESTARTSYS ||
+			    err == -EAGAIN)
+				return VM_FAULT_NOPAGE;
+
+			pr_debug("TTM fault hit %pe.\n", ERR_PTR(err));
+			return VM_FAULT_SIGBUS;
+		}
 	} else {
 		/* Iomem should not be marked encrypted */
 		prot = pgprot_decrypted(prot);
-- 
2.34.1



WARNING: multiple messages have this Message-ID (diff)
From: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>
To: dri-devel@lists.freedesktop.org
Cc: "Miaohe Lin" <linmiaohe@huawei.com>,
	"Thomas Hellström" <thomas.hellstrom@linux.intel.com>,
	"David Hildenbrand" <david@redhat.com>, NeilBrown <neilb@suse.de>,
	"Daniel Vetter" <daniel.vetter@ffwll.ch>,
	intel-gfx@lists.freedesktop.org,
	"Matthew Wilcox (Oracle)" <willy@infradead.org>,
	linux-mm@kvack.org, "Dave Hansen" <dave.hansen@intel.com>,
	linux-graphics-maintainer@vmware.com,
	"Peter Xu" <peterx@redhat.com>,
	"Johannes Weiner" <hannes@cmpxchg.org>,
	"Dave Airlie" <airlied@redhat.com>,
	"Andrew Morton" <akpm@linux-foundation.org>,
	"Christian Koenig" <christian.koenig@amd.com>,
	"Matthew Auld" <matthew.auld@intel.com>
Subject: [RFC PATCH 13/16] drm/ttm: Make the call to ttm_tt_populate() interruptible when faulting
Date: Wed, 15 Feb 2023 17:14:02 +0100	[thread overview]
Message-ID: <20230215161405.187368-14-thomas.hellstrom@linux.intel.com> (raw)
In-Reply-To: <20230215161405.187368-1-thomas.hellstrom@linux.intel.com>

When swapping in, or under memory pressure ttm_tt_populate() may sleep
for a substantiable amount of time. Allow interrupts during the sleep.
This will also allow us to inject -EINTR errors during swapin in upcoming
patches.

Also avoid returning VM_FAULT_OOM, since that will confuse the core
mm, making it print out a confused message and retrying the fault.
Return VM_FAULT_SIGBUS also under OOM conditions.

Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
---
 drivers/gpu/drm/ttm/ttm_bo_vm.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 3ecda6db24b8..80f106bfe385 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -218,14 +218,21 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
 	prot = ttm_io_prot(bo, bo->resource, prot);
 	if (!bo->resource->bus.is_iomem) {
 		struct ttm_operation_ctx ctx = {
-			.interruptible = false,
+			.interruptible = true,
 			.no_wait_gpu = false,
 			.force_alloc = true
 		};
 
 		ttm = bo->ttm;
-		if (ttm_tt_populate(bdev, bo->ttm, &ctx))
-			return VM_FAULT_OOM;
+		err = ttm_tt_populate(bdev, bo->ttm, &ctx);
+		if (err) {
+			if (err == -EINTR || err == -ERESTARTSYS ||
+			    err == -EAGAIN)
+				return VM_FAULT_NOPAGE;
+
+			pr_debug("TTM fault hit %pe.\n", ERR_PTR(err));
+			return VM_FAULT_SIGBUS;
+		}
 	} else {
 		/* Iomem should not be marked encrypted */
 		prot = pgprot_decrypted(prot);
-- 
2.34.1


WARNING: multiple messages have this Message-ID (diff)
From: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>
To: dri-devel@lists.freedesktop.org
Cc: "Miaohe Lin" <linmiaohe@huawei.com>,
	"Thomas Hellström" <thomas.hellstrom@linux.intel.com>,
	"David Hildenbrand" <david@redhat.com>, NeilBrown <neilb@suse.de>,
	"Daniel Vetter" <daniel.vetter@ffwll.ch>,
	intel-gfx@lists.freedesktop.org,
	"Matthew Wilcox (Oracle)" <willy@infradead.org>,
	linux-mm@kvack.org, "Dave Hansen" <dave.hansen@intel.com>,
	linux-graphics-maintainer@vmware.com,
	"Peter Xu" <peterx@redhat.com>,
	"Johannes Weiner" <hannes@cmpxchg.org>,
	"Dave Airlie" <airlied@redhat.com>,
	"Andrew Morton" <akpm@linux-foundation.org>,
	"Christian Koenig" <christian.koenig@amd.com>,
	"Matthew Auld" <matthew.auld@intel.com>
Subject: [Intel-gfx] [RFC PATCH 13/16] drm/ttm: Make the call to ttm_tt_populate() interruptible when faulting
Date: Wed, 15 Feb 2023 17:14:02 +0100	[thread overview]
Message-ID: <20230215161405.187368-14-thomas.hellstrom@linux.intel.com> (raw)
In-Reply-To: <20230215161405.187368-1-thomas.hellstrom@linux.intel.com>

When swapping in, or under memory pressure ttm_tt_populate() may sleep
for a substantiable amount of time. Allow interrupts during the sleep.
This will also allow us to inject -EINTR errors during swapin in upcoming
patches.

Also avoid returning VM_FAULT_OOM, since that will confuse the core
mm, making it print out a confused message and retrying the fault.
Return VM_FAULT_SIGBUS also under OOM conditions.

Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
---
 drivers/gpu/drm/ttm/ttm_bo_vm.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 3ecda6db24b8..80f106bfe385 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -218,14 +218,21 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
 	prot = ttm_io_prot(bo, bo->resource, prot);
 	if (!bo->resource->bus.is_iomem) {
 		struct ttm_operation_ctx ctx = {
-			.interruptible = false,
+			.interruptible = true,
 			.no_wait_gpu = false,
 			.force_alloc = true
 		};
 
 		ttm = bo->ttm;
-		if (ttm_tt_populate(bdev, bo->ttm, &ctx))
-			return VM_FAULT_OOM;
+		err = ttm_tt_populate(bdev, bo->ttm, &ctx);
+		if (err) {
+			if (err == -EINTR || err == -ERESTARTSYS ||
+			    err == -EAGAIN)
+				return VM_FAULT_NOPAGE;
+
+			pr_debug("TTM fault hit %pe.\n", ERR_PTR(err));
+			return VM_FAULT_SIGBUS;
+		}
 	} else {
 		/* Iomem should not be marked encrypted */
 		prot = pgprot_decrypted(prot);
-- 
2.34.1


  parent reply	other threads:[~2023-02-15 16:15 UTC|newest]

Thread overview: 104+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-02-15 16:13 [RFC PATCH 00/16] Add a TTM shrinker Thomas Hellström
2023-02-15 16:13 ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13 ` Thomas Hellström
2023-02-15 16:13 ` [RFC PATCH 01/16] drm/ttm: Fix a NULL pointer dereference Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 17:25   ` Christian König
2023-02-15 17:25     ` [Intel-gfx] " Christian König
2023-02-15 17:25     ` Christian König
2023-02-15 16:13 ` [RFC PATCH 02/16] drm/ttm/pool: Fix ttm_pool_alloc error path Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 17:31   ` Christian König
2023-02-15 17:31     ` [Intel-gfx] " Christian König
2023-02-15 17:31     ` Christian König
2023-02-15 18:02     ` Thomas Hellström
2023-02-15 18:02       ` [Intel-gfx] " Thomas Hellström
2023-02-15 18:02       ` Thomas Hellström
2023-02-15 18:26       ` Christian König
2023-02-15 18:26         ` [Intel-gfx] " Christian König
2023-02-15 18:26         ` Christian König
2023-02-15 18:51         ` Thomas Hellström
2023-02-15 18:51           ` [Intel-gfx] " Thomas Hellström
2023-02-15 18:51           ` Thomas Hellström
2023-02-15 16:13 ` [RFC PATCH 03/16] drm/ttm: Use the BIT macro for the TTM_TT_FLAGs Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 17:33   ` Christian König
2023-02-15 17:33     ` [Intel-gfx] " Christian König
2023-02-15 17:33     ` Christian König
2023-02-15 16:13 ` [RFC PATCH 04/16] drm/ttm, drm/vmwgfx: Update the TTM swapout interface Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 17:39   ` Christian König
2023-02-15 17:39     ` [Intel-gfx] " Christian König
2023-02-15 17:39     ` Christian König
2023-02-15 18:19     ` Thomas Hellström
2023-02-15 18:19       ` [Intel-gfx] " Thomas Hellström
2023-02-15 18:19       ` Thomas Hellström
2023-02-15 18:32       ` Christian König
2023-02-15 18:32         ` [Intel-gfx] " Christian König
2023-02-15 18:32         ` Christian König
2023-02-15 16:13 ` [RFC PATCH 05/16] drm/ttm: Unexport ttm_global_swapout() Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 16:13 ` [RFC PATCH 06/16] drm/ttm: Don't use watermark accounting on shrinkable pools Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 16:13 ` [RFC PATCH 07/16] drm/ttm: Reduce the number of used allocation orders for TTM pages Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 17:42   ` Christian König
2023-02-15 17:42     ` [Intel-gfx] " Christian König
2023-02-15 17:42     ` Christian König
2023-02-15 18:12     ` Thomas Hellström
2023-02-15 18:12       ` [Intel-gfx] " Thomas Hellström
2023-02-15 18:12       ` Thomas Hellström
2023-02-15 18:30       ` Christian König
2023-02-15 18:30         ` [Intel-gfx] " Christian König
2023-02-15 18:30         ` Christian König
2023-02-15 19:00         ` Thomas Hellström
2023-02-15 19:00           ` [Intel-gfx] " Thomas Hellström
2023-02-15 19:00           ` Thomas Hellström
2023-02-16  7:11           ` Christian König
2023-02-16  7:11             ` [Intel-gfx] " Christian König
2023-02-16  7:11             ` Christian König
2023-02-16  7:24             ` Thomas Hellström
2023-02-16  7:24               ` [Intel-gfx] " Thomas Hellström
2023-02-16  7:24               ` Thomas Hellström
2023-02-15 18:15   ` kernel test robot
2023-02-15 20:07   ` kernel test robot
2023-02-15 16:13 ` [Intel-gfx] [RFC PATCH 08/16] drm/ttm: Add a shrinker and shrinker accounting Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 16:13 ` [RFC PATCH 09/16] drm/ttm: Introduce shrink throttling Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 16:13 ` [RFC PATCH 10/16] drm/ttm: Remove pinned bos from shrinkable accounting Thomas Hellström
2023-02-15 16:13   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:13   ` Thomas Hellström
2023-02-15 16:14 ` [RFC PATCH 11/16] drm/ttm: Add a simple api to set / clear purgeable ttm_tt content Thomas Hellström
2023-02-15 16:14   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:14   ` Thomas Hellström
2023-02-15 16:14 ` [RFC PATCH 12/16] mm: Add interfaces to back up and recover folio contents using swap Thomas Hellström
2023-02-15 16:14   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:14   ` Thomas Hellström
2023-02-15 16:14 ` Thomas Hellström [this message]
2023-02-15 16:14   ` [Intel-gfx] [RFC PATCH 13/16] drm/ttm: Make the call to ttm_tt_populate() interruptible when faulting Thomas Hellström
2023-02-15 16:14   ` Thomas Hellström
2023-02-15 16:14 ` [RFC PATCH 14/16] drm/ttm: Provide helpers for shrinking Thomas Hellström
2023-02-15 16:14   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:14   ` Thomas Hellström
2023-02-15 22:00   ` [Intel-gfx] " kernel test robot
2023-02-16  5:41   ` kernel test robot
2023-02-16 16:23   ` kernel test robot
2023-02-15 16:14 ` [RFC PATCH 15/16] drm/ttm: Use fault-injection to test error paths Thomas Hellström
2023-02-15 16:14   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:14   ` Thomas Hellström
2023-02-15 16:14 ` [RFC PATCH 16/16] drm/i915, drm/ttm: Use the TTM shrinker rather than the external shmem pool Thomas Hellström
2023-02-15 16:14   ` [Intel-gfx] " Thomas Hellström
2023-02-15 16:14   ` Thomas Hellström
2023-02-15 19:31 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for Add a TTM shrinker Patchwork
2023-02-15 19:58 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2023-02-16 15:34 ` [Intel-gfx] ✗ Fi.CI.IGT: failure " Patchwork

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230215161405.187368-14-thomas.hellstrom@linux.intel.com \
    --to=thomas.hellstrom@linux.intel.com \
    --cc=airlied@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=christian.koenig@amd.com \
    --cc=daniel.vetter@ffwll.ch \
    --cc=dave.hansen@intel.com \
    --cc=david@redhat.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=hannes@cmpxchg.org \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=linmiaohe@huawei.com \
    --cc=linux-graphics-maintainer@vmware.com \
    --cc=linux-mm@kvack.org \
    --cc=matthew.auld@intel.com \
    --cc=neilb@suse.de \
    --cc=peterx@redhat.com \
    --cc=willy@infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.