From: "Sharma, Nishit" <nishit.sharma@intel.com>
To: Arvind Yadav <arvind.yadav@intel.com>, <igt-dev@lists.freedesktop.org>
Cc: <matthew.brost@intel.com>, <himal.prasad.ghimiray@intel.com>,
<thomas.hellstrom@linux.intel.com>, <pravalika.gurram@intel.com>
Subject: Re: [PATCH i-g-t v6 4/9] tests/intel/xe_madvise: Add purged-mmap-blocked subtest
Date: Mon, 6 Apr 2026 18:04:16 +0530 [thread overview]
Message-ID: <41012de6-8a70-49ed-9332-1746598c8e51@intel.com> (raw)
In-Reply-To: <20260325124426.3265234-5-arvind.yadav@intel.com>
On 3/25/2026 6:14 PM, Arvind Yadav wrote:
> Add a new purged-mmap-blocked subtest that triggers an actual purge
> via memory pressure and verifies that mmap() fails with -EINVAL once
> the BO backing store has been permanently discarded.
>
> The purgeable check moved from xe_gem_mmap_offset_ioctl()
> into a new xe_gem_object_mmap() callback, so the blocking point is now
> mmap() itself rather than the mmap offset ioctl:
>
> - DRM_IOCTL_XE_GEM_MMAP_OFFSET: always succeeds regardless of
> purgeable state (just returns the pre-allocated offset)
> - mmap() on DONTNEED BO: fails with -EBUSY (temporary state)
> - mmap() on purged BO: fails with -EINVAL (permanent, no backing store)
>
> v5:
> - Add purged-mmap-blocked subtest to verify mmap is blocked after
> BO backing store is permanently purged.
>
> v6:
> - DRM_IOCTL_XE_GEM_MMAP_OFFSET always succeeds; the purgeable check
> now happens in xe_gem_object_mmap() at mmap() time. For purged BOs,
> assert mmap() fails with -EINVAL.
>
> Cc: Nishit Sharma <nishit.sharma@intel.com>
> Cc: Matthew Brost <matthew.brost@intel.com>
> Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
> Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
> Cc: Pravalika Gurram <pravalika.gurram@intel.com>
> Signed-off-by: Arvind Yadav <arvind.yadav@intel.com>
> ---
> tests/intel/xe_madvise.c | 42 ++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 42 insertions(+)
>
> diff --git a/tests/intel/xe_madvise.c b/tests/intel/xe_madvise.c
> index de4a6e34c..81e05b6d4 100644
> --- a/tests/intel/xe_madvise.c
> +++ b/tests/intel/xe_madvise.c
> @@ -170,6 +170,42 @@ static void purgeable_setup_simple_bo(int fd, uint32_t *vm, uint32_t *bo,
> xe_wait_ufence(fd, &sync_val, 1, 0, NSEC_PER_SEC);
> }
>
> +/**
> + * SUBTEST: purged-mmap-blocked
> + * Description: After BO is purged, verify mmap() fails with -EINVAL
> + * Test category: functionality test
> + */
> +static void test_purged_mmap_blocked(int fd)
> +{
> + uint32_t bo, vm;
> + uint64_t addr = PURGEABLE_ADDR;
> + size_t bo_size = PURGEABLE_BO_SIZE;
> + struct drm_xe_gem_mmap_offset mmo = {};
> + void *ptr;
> +
> + purgeable_setup_simple_bo(fd, &vm, &bo, addr, bo_size, false);
> + if (!purgeable_mark_and_verify_purged(fd, vm, addr, bo_size)) {
This function should be defined in this patch. Going through
implementation available in Patch-3/9. In that below code used
vram_size = xe_visible_vram_size(fd, 0);
What it returns in igfx platforms? Instead we can use
/* Purgeable BO either in VRAM or SMEM */
mem_size = xe_has_vram(pf_fd)?xe_visible_vram_size(fd,
0):igt_get_total_ram_mb() << 20;
if (vram_size > 0) {
/* dGPU: pressure VRAM to trigger purgeable reclaim */
mem_size = vram_size;
} else {
/*
* iGPU: purgeable BOs reside in system memory. Use *total*
* RAM (not just available) as the baseline so that we
always
* over-commit regardless of how much is already in use.
*/
mem_size = igt_get_total_ram_mb() << 20;
}
> + gem_close(fd, bo);
> + xe_vm_destroy(fd, vm);
> + igt_skip("Unable to induce purge on this platform/config");
> + }
static bool purgeable_mark_and_verify_purged(int fd, uint32_t vm,
uint64_t addr, size_t size) --> this should be part of this patch
{
uint32_t retained;
/* Mark as DONTNEED */
retained = xe_vm_madvise_purgeable(fd, vm, addr, size,
DRM_XE_VMA_PURGEABLE_STATE_DONTNEED);
if (retained != 1)
So here if BO was purged after above call it'll return 0, so if (0 != 1)
will be true and it'll return false.
return false;
/* Trigger memory pressure */
trigger_memory_pressure(fd, vm);
/* Verify purged */
retained = xe_vm_madvise_purgeable(fd, vm, addr, size,
DRM_XE_VMA_PURGEABLE_STATE_WILLNEED);
return retained == 0;
}
static void test_purged_mmap_blocked(int fd)
{
uint32_t bo, vm;
uint64_t addr = PURGEABLE_ADDR;
size_t bo_size = PURGEABLE_BO_SIZE;
struct drm_xe_gem_mmap_offset mmo = {};
void *ptr;
purgeable_setup_simple_bo(fd, &vm, &bo, addr, bo_size, false);
if (!purgeable_mark_and_verify_purged(fd, vm, addr, bo_size)) {
Here if returned false from above which is basically if BO was purged
retained = 0 then skip is called which is not the intention
gem_close(fd, bo);
xe_vm_destroy(fd, vm);
igt_skip("Unable to induce purge on this platform/config");
}
/*
* Getting the mmap offset is always allowed regardless of
purgeable
* state - the blocking happens at mmap() time
(xe_gem_object_mmap).
* For a purged BO, mmap() must fail with -EINVAL (no backing
store).
*/
mmo.handle = bo;
igt_assert_eq(igt_ioctl(fd, DRM_IOCTL_XE_GEM_MMAP_OFFSET,
&mmo), 0);
ptr = mmap(NULL, bo_size, PROT_READ | PROT_WRITE, MAP_SHARED,
fd, mmo.offset);
igt_assert_eq_u64((uint64_t)ptr, (uint64_t)MAP_FAILED);
igt_assert_eq(errno, EINVAL);
gem_close(fd, bo);
xe_vm_destroy(fd, vm);
}
> +
> + /*
> + * Getting the mmap offset is always allowed regardless of purgeable
> + * state - the blocking happens at mmap() time (xe_gem_object_mmap).
> + * For a purged BO, mmap() must fail with -EINVAL (no backing store).
> + */
> + mmo.handle = bo;
> + igt_assert_eq(igt_ioctl(fd, DRM_IOCTL_XE_GEM_MMAP_OFFSET, &mmo), 0);
> +
> + ptr = mmap(NULL, bo_size, PROT_READ | PROT_WRITE, MAP_SHARED, fd, mmo.offset);
> + igt_assert_eq_u64((uint64_t)ptr, (uint64_t)MAP_FAILED);
> + igt_assert_eq(errno, EINVAL);
> +
> + gem_close(fd, bo);
> + xe_vm_destroy(fd, vm);
> +}
> +
> /**
> * SUBTEST: dontneed-before-mmap
> * Description: Mark BO as DONTNEED before mmap, verify mmap() fails with -EBUSY
> @@ -226,6 +262,12 @@ int igt_main()
> break;
> }
>
> + igt_subtest("purged-mmap-blocked")
> + xe_for_each_engine(fd, hwe) {
> + test_purged_mmap_blocked(fd);
> + break;
> + }
> +
> igt_fixture() {
> xe_device_put(fd);
> drm_close_driver(fd);
next prev parent reply other threads:[~2026-04-06 12:34 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-25 12:44 [PATCH i-g-t v6 0/9] tests/xe: Add purgeable memory madvise tests for system allocator Arvind Yadav
2026-03-25 12:44 ` [PATCH i-g-t v6 1/9] drm-uapi/xe_drm: Add UAPI support for purgeable buffer objects Arvind Yadav
2026-04-06 5:23 ` Sharma, Nishit
2026-04-06 6:00 ` Yadav, Arvind
2026-03-25 12:44 ` [PATCH i-g-t v6 2/9] lib/xe: Add purgeable memory ioctl support Arvind Yadav
2026-04-06 6:59 ` Sharma, Nishit
2026-04-07 3:18 ` Yadav, Arvind
2026-03-25 12:44 ` [PATCH i-g-t v6 3/9] tests/intel/xe_madvise: Add dontneed-before-mmap subtest Arvind Yadav
2026-04-06 9:53 ` Sharma, Nishit
2026-04-06 10:30 ` Sharma, Nishit
2026-04-07 4:21 ` Yadav, Arvind
2026-03-25 12:44 ` [PATCH i-g-t v6 4/9] tests/intel/xe_madvise: Add purged-mmap-blocked subtest Arvind Yadav
2026-04-06 12:34 ` Sharma, Nishit [this message]
2026-04-07 5:09 ` Yadav, Arvind
2026-03-25 12:44 ` [PATCH i-g-t v6 5/9] tests/intel/xe_madvise: Add dontneed-after-mmap subtest Arvind Yadav
2026-04-06 13:33 ` Sharma, Nishit
2026-04-07 5:15 ` Yadav, Arvind
2026-03-25 12:44 ` [PATCH i-g-t v6 6/9] tests/intel/xe_madvise: Add dontneed-before-exec subtest Arvind Yadav
2026-04-06 16:48 ` Sharma, Nishit
2026-04-07 5:29 ` Yadav, Arvind
2026-03-25 12:44 ` [PATCH i-g-t v6 7/9] tests/intel/xe_madvise: Add dontneed-after-exec subtest Arvind Yadav
2026-04-07 14:51 ` Sharma, Nishit
2026-03-25 12:44 ` [PATCH i-g-t v6 8/9] tests/intel/xe_madvise: Add per-vma-tracking subtest Arvind Yadav
2026-04-07 7:20 ` Sharma, Nishit
2026-04-07 8:49 ` Yadav, Arvind
2026-03-25 12:44 ` [PATCH i-g-t v6 9/9] tests/intel/xe_madvise: Add per-vma-protection subtest Arvind Yadav
2026-04-07 7:31 ` Sharma, Nishit
2026-04-07 8:54 ` Yadav, Arvind
2026-03-25 22:59 ` ✓ Xe.CI.BAT: success for tests/xe: Add purgeable memory madvise tests for system allocator (rev6) Patchwork
2026-03-25 23:15 ` ✓ i915.CI.BAT: " Patchwork
2026-03-26 9:19 ` ✗ Xe.CI.FULL: failure " Patchwork
2026-03-26 11:22 ` ✓ i915.CI.Full: success " 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=41012de6-8a70-49ed-9332-1746598c8e51@intel.com \
--to=nishit.sharma@intel.com \
--cc=arvind.yadav@intel.com \
--cc=himal.prasad.ghimiray@intel.com \
--cc=igt-dev@lists.freedesktop.org \
--cc=matthew.brost@intel.com \
--cc=pravalika.gurram@intel.com \
--cc=thomas.hellstrom@linux.intel.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox