All of lore.kernel.org
 help / color / mirror / Atom feed
* [merged mm-stable] lib-test_hmm-check-alloc_page_vma-return-value-and-handle-oom.patch removed from -mm tree
@ 2026-06-02 22:25 Andrew Morton
  0 siblings, 0 replies; 2+ messages in thread
From: Andrew Morton @ 2026-06-02 22:25 UTC (permalink / raw)
  To: mm-commits, leon, jgg, apopple, liuqiang, akpm


The quilt patch titled
     Subject: lib/test_hmm: check alloc_page_vma() return value and handle OOM
has been removed from the -mm tree.  Its filename was
     lib-test_hmm-check-alloc_page_vma-return-value-and-handle-oom.patch

This patch was dropped because it was merged into the mm-stable branch
of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

------------------------------------------------------
From: Qiang Liu <liuqiang@kylinos.cn>
Subject: lib/test_hmm: check alloc_page_vma() return value and handle OOM
Date: Thu, 21 May 2026 10:18:58 +0800

Check alloc_page_vma() return status for page allocation failures, free
allocated pages and return VM_FAULT_OOM on error.

Handle return codes of dmirror_devmem_fault_alloc_and_copy(), call
migrate_vma_finalize() to remove migration entries from
migrate_vma_setup().

Link: https://lore.kernel.org/20260521021858.21511-1-liuqiangneo@163.com
Signed-off-by: Qiang Liu <liuqiang@kylinos.cn>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 lib/test_hmm.c |   29 +++++++++++++++++++++++++++--
 1 file changed, 27 insertions(+), 2 deletions(-)

--- a/lib/test_hmm.c~lib-test_hmm-check-alloc_page_vma-return-value-and-handle-oom
+++ a/lib/test_hmm.c
@@ -1063,6 +1063,25 @@ static vm_fault_t dmirror_devmem_fault_a
 			/* Try with smaller pages if large allocation fails */
 			if (!dpage && order) {
 				dpage = alloc_page_vma(GFP_HIGHUSER_MOVABLE, args->vma, addr);
+				if (!dpage) {
+					/* Unlock and free pages already allocated. */
+					while (i > 0) {
+						struct page *fpage;
+
+						fpage = migrate_pfn_to_page(dst[--i]);
+						unlock_page(fpage);
+						__free_page(fpage);
+					}
+					/* Clear remaining dst entries to avoid
+					 * migrate_vma_pages/finalize() using
+					 * uninitialized values.
+					 */
+					while (i < (1 << order)) {
+						dst[i] = 0;
+						i++;
+					}
+					return VM_FAULT_OOM;
+				}
 				lock_page(dpage);
 				dst[i] = migrate_pfn(page_to_pfn(dpage));
 				dst_page = pfn_to_page(page_to_pfn(dpage));
@@ -1148,7 +1167,11 @@ static int dmirror_migrate_to_system(str
 			goto out;
 
 		pr_debug("Migrating from device mem to sys mem\n");
-		dmirror_devmem_fault_alloc_and_copy(&args, dmirror);
+		ret = dmirror_devmem_fault_alloc_and_copy(&args, dmirror);
+		if (ret) {
+			migrate_vma_finalize(&args);
+			goto out;
+		}
 
 		migrate_vma_pages(&args);
 		cmd->cpages += dmirror_successful_migrated_pages(&args);
@@ -1689,8 +1712,10 @@ static vm_fault_t dmirror_devmem_fault(s
 	}
 
 	ret = dmirror_devmem_fault_alloc_and_copy(&args, dmirror);
-	if (ret)
+	if (ret) {
+		migrate_vma_finalize(&args);
 		goto err;
+	}
 	migrate_vma_pages(&args);
 	/*
 	 * No device finalize step is needed since
_

Patches currently in -mm which might be from liuqiang@kylinos.cn are



^ permalink raw reply	[flat|nested] 2+ messages in thread

* [merged mm-stable] lib-test_hmm-check-alloc_page_vma-return-value-and-handle-oom.patch removed from -mm tree
@ 2026-06-09  1:22 Andrew Morton
  0 siblings, 0 replies; 2+ messages in thread
From: Andrew Morton @ 2026-06-09  1:22 UTC (permalink / raw)
  To: mm-commits, leon, jgg, apopple, liuqiang, akpm


The quilt patch titled
     Subject: lib/test_hmm: check alloc_page_vma() return value and handle OOM
has been removed from the -mm tree.  Its filename was
     lib-test_hmm-check-alloc_page_vma-return-value-and-handle-oom.patch

This patch was dropped because it was merged into the mm-stable branch
of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

------------------------------------------------------
From: Qiang Liu <liuqiang@kylinos.cn>
Subject: lib/test_hmm: check alloc_page_vma() return value and handle OOM
Date: Thu, 21 May 2026 10:18:58 +0800

Check alloc_page_vma() return status for page allocation failures, free
allocated pages and return VM_FAULT_OOM on error.

Handle return codes of dmirror_devmem_fault_alloc_and_copy(), call
migrate_vma_finalize() to remove migration entries from
migrate_vma_setup().

Link: https://lore.kernel.org/20260521021858.21511-1-liuqiangneo@163.com
Signed-off-by: Qiang Liu <liuqiang@kylinos.cn>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Leon Romanovsky <leon@kernel.org>
[akpm@linux-foundation.org: fix dmirror_devmem_fault_alloc_and_copy() retval handling]
  Link: https://lore.kernel.org/oe-kbuild-all/202606011329.zWs2BKy4-lkp@intel.com/
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 lib/test_hmm.c |   29 +++++++++++++++++++++++++++--
 1 file changed, 27 insertions(+), 2 deletions(-)

--- a/lib/test_hmm.c~lib-test_hmm-check-alloc_page_vma-return-value-and-handle-oom
+++ a/lib/test_hmm.c
@@ -1063,6 +1063,25 @@ static vm_fault_t dmirror_devmem_fault_a
 			/* Try with smaller pages if large allocation fails */
 			if (!dpage && order) {
 				dpage = alloc_page_vma(GFP_HIGHUSER_MOVABLE, args->vma, addr);
+				if (!dpage) {
+					/* Unlock and free pages already allocated. */
+					while (i > 0) {
+						struct page *fpage;
+
+						fpage = migrate_pfn_to_page(dst[--i]);
+						unlock_page(fpage);
+						__free_page(fpage);
+					}
+					/* Clear remaining dst entries to avoid
+					 * migrate_vma_pages/finalize() using
+					 * uninitialized values.
+					 */
+					while (i < (1 << order)) {
+						dst[i] = 0;
+						i++;
+					}
+					return VM_FAULT_OOM;
+				}
 				lock_page(dpage);
 				dst[i] = migrate_pfn(page_to_pfn(dpage));
 				dst_page = pfn_to_page(page_to_pfn(dpage));
@@ -1148,7 +1167,11 @@ static int dmirror_migrate_to_system(str
 			goto out;
 
 		pr_debug("Migrating from device mem to sys mem\n");
-		dmirror_devmem_fault_alloc_and_copy(&args, dmirror);
+		if (dmirror_devmem_fault_alloc_and_copy(&args, dmirror)) {
+			migrate_vma_finalize(&args);
+			ret = -ENOMEM;
+			goto out;
+		}
 
 		migrate_vma_pages(&args);
 		cmd->cpages += dmirror_successful_migrated_pages(&args);
@@ -1689,8 +1712,10 @@ static vm_fault_t dmirror_devmem_fault(s
 	}
 
 	ret = dmirror_devmem_fault_alloc_and_copy(&args, dmirror);
-	if (ret)
+	if (ret) {
+		migrate_vma_finalize(&args);
 		goto err;
+	}
 	migrate_vma_pages(&args);
 	/*
 	 * No device finalize step is needed since
_

Patches currently in -mm which might be from liuqiang@kylinos.cn are



^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2026-06-09  1:22 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-06-02 22:25 [merged mm-stable] lib-test_hmm-check-alloc_page_vma-return-value-and-handle-oom.patch removed from -mm tree Andrew Morton
  -- strict thread matches above, loose matches on Subject: below --
2026-06-09  1:22 Andrew Morton

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.