1. Patchew
  2. linux
  3. [PATCH 00/13] Enable compound page for p2pdma memory
  4. View patch

[PATCH 02/13] PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()

Hou Tao posted 13 patches 1 month, 3 weeks ago
[PATCH 02/13] PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()
Posted by Hou Tao 1 month, 3 weeks ago 
From: Hou Tao <houtao1@huawei.com>

Commit b7e282378773 has already changed the initial page refcount of
p2pdma page from one to zero, however, in p2pmem_alloc_mmap() it uses
"VM_WARN_ON_ONCE_PAGE(!page_ref_count(page))" to assert the initial page
refcount should not be zero and the following will be reported when
CONFIG_DEBUG_VM is enabled:

 page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x380400000
 flags: 0x20000000002000(reserved|node=0|zone=4)
 raw: 0020000000002000 ff1100015e3ab440 0000000000000000 0000000000000000
 raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
 page dumped because: VM_WARN_ON_ONCE_PAGE(!page_ref_count(page))
 ------------[ cut here ]------------
 WARNING: CPU: 5 PID: 449 at drivers/pci/p2pdma.c:240 p2pmem_alloc_mmap+0x83a/0xa60

Fix by using "page_ref_count(page)" as the assertion condition.

Fixes: b7e282378773 ("mm/mm_init: move p2pdma page refcount initialisation to p2pdma")
Signed-off-by: Hou Tao <houtao1@huawei.com>
---
 drivers/pci/p2pdma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/pci/p2pdma.c b/drivers/pci/p2pdma.c
index 218c1f5252b6..dd64ec830fdd 100644
--- a/drivers/pci/p2pdma.c
+++ b/drivers/pci/p2pdma.c
@@ -147,7 +147,7 @@ static int p2pmem_alloc_mmap(struct file *filp, struct kobject *kobj,
 		 * we have just allocated the page no one else should be
 		 * using it.
 		 */
-		VM_WARN_ON_ONCE_PAGE(!page_ref_count(page), page);
+		VM_WARN_ON_ONCE_PAGE(page_ref_count(page), page);
 		set_page_count(page, 1);
 		ret = vm_insert_page(vma, vaddr, page);
 		if (ret) {
-- 
2.29.2
Re: [PATCH 02/13] PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()
Posted by Alistair Popple 1 month ago 
On 2025-12-20 at 15:04 +1100, Hou Tao <houtao@huaweicloud.com> wrote...
> From: Hou Tao <houtao1@huawei.com>
> 
> Commit b7e282378773 has already changed the initial page refcount of
> p2pdma page from one to zero, however, in p2pmem_alloc_mmap() it uses
> "VM_WARN_ON_ONCE_PAGE(!page_ref_count(page))" to assert the initial page
> refcount should not be zero and the following will be reported when
> CONFIG_DEBUG_VM is enabled:
> 
>  page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x380400000
>  flags: 0x20000000002000(reserved|node=0|zone=4)
>  raw: 0020000000002000 ff1100015e3ab440 0000000000000000 0000000000000000
>  raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
>  page dumped because: VM_WARN_ON_ONCE_PAGE(!page_ref_count(page))
>  ------------[ cut here ]------------
>  WARNING: CPU: 5 PID: 449 at drivers/pci/p2pdma.c:240 p2pmem_alloc_mmap+0x83a/0xa60
> 
> Fix by using "page_ref_count(page)" as the assertion condition.
> 
> Fixes: b7e282378773 ("mm/mm_init: move p2pdma page refcount initialisation to p2pdma")

Argh, thanks for fixing that.

Reviewed-by: Alistair Popple <apopple@nvidia.com>

> Signed-off-by: Hou Tao <houtao1@huawei.com>
> ---
>  drivers/pci/p2pdma.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/p2pdma.c b/drivers/pci/p2pdma.c
> index 218c1f5252b6..dd64ec830fdd 100644
> --- a/drivers/pci/p2pdma.c
> +++ b/drivers/pci/p2pdma.c
> @@ -147,7 +147,7 @@ static int p2pmem_alloc_mmap(struct file *filp, struct kobject *kobj,
>  		 * we have just allocated the page no one else should be
>  		 * using it.
>  		 */
> -		VM_WARN_ON_ONCE_PAGE(!page_ref_count(page), page);
> +		VM_WARN_ON_ONCE_PAGE(page_ref_count(page), page);
>  		set_page_count(page, 1);
>  		ret = vm_insert_page(vma, vaddr, page);
>  		if (ret) {
> -- 
> 2.29.2
>
Re: [PATCH 02/13] PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()
Posted by Logan Gunthorpe 1 month, 2 weeks ago 

On 2025-12-19 21:04, Hou Tao wrote:
> From: Hou Tao <houtao1@huawei.com>
> 
> Commit b7e282378773 has already changed the initial page refcount of
> p2pdma page from one to zero, however, in p2pmem_alloc_mmap() it uses
> "VM_WARN_ON_ONCE_PAGE(!page_ref_count(page))" to assert the initial page
> refcount should not be zero and the following will be reported when
> CONFIG_DEBUG_VM is enabled:
> 
>  page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x380400000
>  flags: 0x20000000002000(reserved|node=0|zone=4)
>  raw: 0020000000002000 ff1100015e3ab440 0000000000000000 0000000000000000
>  raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
>  page dumped because: VM_WARN_ON_ONCE_PAGE(!page_ref_count(page))
>  ------------[ cut here ]------------
>  WARNING: CPU: 5 PID: 449 at drivers/pci/p2pdma.c:240 p2pmem_alloc_mmap+0x83a/0xa60
> 
> Fix by using "page_ref_count(page)" as the assertion condition.
> 
> Fixes: b7e282378773 ("mm/mm_init: move p2pdma page refcount initialisation to p2pdma")
> Signed-off-by: Hou Tao <houtao1@huawei.com>

Thanks for the fix

Reviewed-by: Logan Gunthorpe <logang@deltatee.com>

Logan
Re: [PATCH 02/13] PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()
Posted by Christoph Hellwig 1 month ago 
On Mon, Dec 22, 2025 at 09:50:18AM -0700, Logan Gunthorpe wrote:
> > Fixes: b7e282378773 ("mm/mm_init: move p2pdma page refcount initialisation to p2pdma")
> > Signed-off-by: Hou Tao <houtao1@huawei.com>
> 
> Thanks for the fix
> 
> Reviewed-by: Logan Gunthorpe <logang@deltatee.com>

Can we get these fixes queued up ASAP while discussing the new
feature?
Re: [PATCH 02/13] PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()
Posted by Bjorn Helgaas 1 month ago 
On Wed, Jan 07, 2026 at 03:39:42PM +0100, Christoph Hellwig wrote:
> On Mon, Dec 22, 2025 at 09:50:18AM -0700, Logan Gunthorpe wrote:
> > > Fixes: b7e282378773 ("mm/mm_init: move p2pdma page refcount initialisation to p2pdma")
> > > Signed-off-by: Hou Tao <houtao1@huawei.com>
> > 
> > Thanks for the fix
> > 
> > Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
> 
> Can we get these fixes queued up ASAP while discussing the new
> feature?

I assume you mean the first two patches:

  PCI/P2PDMA: Release the per-cpu ref of pgmap when vm_insert_page() fails
  PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()

I can certainly queue these up.  The second is a warning fix, and the
first fixes a hang that has apparently been around since v6.2
(7e9c7ef83d78 ("PCI/P2PDMA: Allow userspace VMA allocations through
sysfs")).

Given that, I would ordinarily target the v6.20 merge window, but the
"ASAP" suggests more urgency.  Do you want one or both for v6.19?

Bjorn
Re: [PATCH 02/13] PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()
Posted by Bjorn Helgaas 1 month ago 
On Wed, Jan 07, 2026 at 11:17:24AM -0600, Bjorn Helgaas wrote:
> On Wed, Jan 07, 2026 at 03:39:42PM +0100, Christoph Hellwig wrote:
> > On Mon, Dec 22, 2025 at 09:50:18AM -0700, Logan Gunthorpe wrote:
> > > > Fixes: b7e282378773 ("mm/mm_init: move p2pdma page refcount initialisation to p2pdma")
> > > > Signed-off-by: Hou Tao <houtao1@huawei.com>
> > > 
> > > Thanks for the fix
> > > 
> > > Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
> > 
> > Can we get these fixes queued up ASAP while discussing the new
> > feature?
> 
> I assume you mean the first two patches:
> 
>   PCI/P2PDMA: Release the per-cpu ref of pgmap when vm_insert_page() fails
>   PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()
> 
> I can certainly queue these up.  The second is a warning fix, and the
> first fixes a hang that has apparently been around since v6.2
> (7e9c7ef83d78 ("PCI/P2PDMA: Allow userspace VMA allocations through
> sysfs")).
> 
> Given that, I would ordinarily target the v6.20 merge window, but the
> "ASAP" suggests more urgency.  Do you want one or both for v6.19?

I put the patches above on pci/p2pdma for v6.20 until I hear
otherwise.

Bjorn
Re: [PATCH 02/13] PCI/P2PDMA: Fix the warning condition in p2pmem_alloc_mmap()
Posted by Christoph Hellwig 1 month ago 
On Wed, Jan 07, 2026 at 02:34:39PM -0600, Bjorn Helgaas wrote:
> > 
> > Given that, I would ordinarily target the v6.20 merge window, but the
> > "ASAP" suggests more urgency.  Do you want one or both for v6.19?
> 
> I put the patches above on pci/p2pdma for v6.20 until I hear
> otherwise.

Sounds good, thanks!

Patchew 2.1-devel-b67e4c2

© 2016 - 2026 Red Hat, Inc.