From: Philip Yang <Philip.Yang@amd.com>
To: <amd-gfx@lists.freedesktop.org>
Cc: <Felix.Kuehling@amd.com>, <christian.koenig@amd.com>,
<david.yatsin@amd.com>, <pierre-eric.pelloux-prayer@amd.com>,
<kent.russell@amd.com>, Philip Yang <Philip.Yang@amd.com>
Subject: [PATCH v5 0/6] drm/amdkfd: Move gfx9 MQD to HBM
Date: Tue, 9 Dec 2025 18:43:22 -0500 [thread overview]
Message-ID: <20251209234328.625670-1-Philip.Yang@amd.com> (raw)
To reduce multiple queues switch latency, move user queues MQD to HBM,
and map on GART with mtype RW.
v5:
- patch 1 update commit message
- patch 5 separate GART alloc helpers
- patch 6 use resource cursor to handle GART entries
v4:
- patch 1 remove the executable bit change, cc stable
- patch 5 move GART helper functions to amdgpu_gtt_mgr.c
- add patch 6 to update MQD GART mapping using resource cursor
v3:
- add patch 1 to fix gfx9 mtype update bug
- patch 2 use ASIC specific mtype
- patch 5 use drm mm to alloc GART entries and store in mqd obj
v2:
- patch 4 GART mapping use MC address, vram_base_offset + physical address
Philip Yang (6):
drm/amdgpu: Fix gfx9 update PTE mtype flag
drm/amdkfd: Bind MQD in GART with mtype RW
drm/amdkfd: Add domain parameter to alloc kernel BO
drm/amdkfd: Move gfx9 MQD to VRAM domain
drm/amdgpu: Add helper to alloc GART entries
drm/amdkfd: Map VRAM MQD on GART
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 13 +--
drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 6 +-
.../gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 2 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_gtt_mgr.c | 27 ++++++
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 96 ++++++++++++++++++-
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h | 11 ++-
drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 3 -
drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 8 +-
drivers/gpu/drm/amd/amdkfd/kfd_debug.c | 3 +-
drivers/gpu/drm/amd/amdkfd/kfd_device.c | 13 +--
.../drm/amd/amdkfd/kfd_device_queue_manager.c | 7 +-
drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c | 16 ++--
.../drm/amd/amdkfd/kfd_mqd_manager_v12_1.c | 4 +-
.../gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c | 18 +++-
drivers/gpu/drm/amd/amdkfd/kfd_priv.h | 3 +-
drivers/gpu/drm/amd/amdkfd/kfd_process.c | 2 +-
.../amd/amdkfd/kfd_process_queue_manager.c | 12 ++-
17 files changed, 195 insertions(+), 49 deletions(-)
--
2.50.1
next reply other threads:[~2025-12-09 23:44 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-12-09 23:43 Philip Yang [this message]
2025-12-09 23:43 ` [PATCH v5 1/6] drm/amdgpu: Fix gfx9 update PTE mtype flag Philip Yang
2025-12-09 23:43 ` [PATCH v5 2/6] drm/amdkfd: Bind MQD in GART with mtype RW Philip Yang
2025-12-09 23:43 ` [PATCH v5 3/6] drm/amdkfd: Add domain parameter to alloc kernel BO Philip Yang
2025-12-09 23:43 ` [PATCH v5 4/6] drm/amdkfd: Move gfx9 MQD to VRAM domain Philip Yang
2025-12-09 23:43 ` [PATCH v5 5/6] drm/amdgpu: Add helper to alloc GART entries Philip Yang
2025-12-10 12:57 ` Pierre-Eric Pelloux-Prayer
2025-12-10 14:05 ` Philip Yang
2025-12-15 15:14 ` Christian König
2025-12-15 15:50 ` Philip Yang
2025-12-09 23:43 ` [PATCH v5 6/6] drm/amdkfd: Map VRAM MQD on GART Philip Yang
2025-12-15 15:20 ` Christian König
2025-12-15 16:35 ` Philip Yang
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=20251209234328.625670-1-Philip.Yang@amd.com \
--to=philip.yang@amd.com \
--cc=Felix.Kuehling@amd.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=christian.koenig@amd.com \
--cc=david.yatsin@amd.com \
--cc=kent.russell@amd.com \
--cc=pierre-eric.pelloux-prayer@amd.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