From: Gustavo Sousa <gustavo.sousa@intel.com>
To: Matt Roper <matthew.d.roper@intel.com>, <intel-xe@lists.freedesktop.org>
Cc: Violet Monti <violet.monti@intel.com>
Subject: Re: [PATCH 3/4] drm/xe: Move engines' LRC programming RTP table off the stack
Date: Tue, 16 Jun 2026 19:26:47 -0300 [thread overview]
Message-ID: <87wlvyt93s.fsf@intel.com> (raw)
In-Reply-To: <20260616195417.GM6214@mdroper-desk1.amr.corp.intel.com>
Matt Roper <matthew.d.roper@intel.com> writes:
> On Tue, Jun 16, 2026 at 12:21:53PM -0700, Matt Roper wrote:
>> The 'lrc_setup' RTP table was allocated on the stack because it wasn't
>> truly constant and needed to calculate the proper value for BLIT_CCTL at
>> runtime based on other stack variables. Using the FIELD_SET_FUNC action
>> allows us to make the table itself truly constant and move it off the
>> stack; the BLIT_CCTL value is now calculated during RTP table
>> processing.
>>
>> Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
>> ---
>> drivers/gpu/drm/xe/xe_hw_engine.c | 60 ++++++++++++++++++++-------------------
>> 1 file changed, 31 insertions(+), 29 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/xe/xe_hw_engine.c b/drivers/gpu/drm/xe/xe_hw_engine.c
>> index 98265293f2dc..0b688b851b71 100644
>> --- a/drivers/gpu/drm/xe/xe_hw_engine.c
>> +++ b/drivers/gpu/drm/xe/xe_hw_engine.c
>> @@ -337,39 +337,41 @@ static bool xe_rtp_cfeg_wmtp_disabled(const struct xe_device *xe,
>> return xe_mmio_read32(&hwe->gt->mmio, XEHP_FUSE4) & CFEG_WMTP_DISABLE;
>> }
>>
>> +static u32 blit_cctl_val(struct xe_gt *gt, struct xe_hw_engine *hwe)
>> +{
>> + return REG_FIELD_PREP(BLIT_CCTL_DST_MOCS_MASK, gt->mocs.uc_index) |
>> + REG_FIELD_PREP(BLIT_CCTL_SRC_MOCS_MASK, gt->mocs.uc_index);
>> +}
>> +
>> +const struct xe_rtp_table_sr lrc_setup = XE_RTP_TABLE_SR(
>
> This (and the table in the next patch) should have been 'static.' I
With the addition of the "static" keyword for lrc_setup,
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
> initially had VISIBLE_IF_KUNIT here, but decided to drop that
> until/unless we actually have kunit tests that operate on them.
>
>
> Matt
>
>> + /*
>> + * Some blitter commands do not have a field for MOCS, those
>> + * commands will use MOCS index pointed by BLIT_CCTL.
>> + * BLIT_CCTL registers are needed to be programmed to un-cached.
>> + */
>> + { XE_RTP_NAME("BLIT_CCTL_default_MOCS"),
>> + XE_RTP_RULES(GRAPHICS_VERSION_RANGE(1200, 1274),
>> + ENGINE_CLASS(COPY)),
>> + XE_RTP_ACTIONS(FIELD_SET_FUNC(BLIT_CCTL(0),
>> + BLIT_CCTL_DST_MOCS_MASK |
>> + BLIT_CCTL_SRC_MOCS_MASK,
>> + blit_cctl_val,
>> + XE_RTP_ACTION_FLAG(ENGINE_BASE)))
>> + },
>> + /* Disable WMTP if HW doesn't support it */
>> + { XE_RTP_NAME("DISABLE_WMTP_ON_UNSUPPORTED_HW"),
>> + XE_RTP_RULES(FUNC(xe_rtp_cfeg_wmtp_disabled)),
>> + XE_RTP_ACTIONS(FIELD_SET(CS_CHICKEN1(0),
>> + PREEMPT_GPGPU_LEVEL_MASK,
>> + PREEMPT_GPGPU_THREAD_GROUP_LEVEL)),
>> + XE_RTP_ENTRY_FLAG(FOREACH_ENGINE)
>> + },
>> +);
>> +
>> static void
>> hw_engine_setup_default_lrc_state(struct xe_hw_engine *hwe)
>> {
>> - struct xe_gt *gt = hwe->gt;
>> - const u8 mocs_write_idx = gt->mocs.uc_index;
>> - const u8 mocs_read_idx = gt->mocs.uc_index;
>> - u32 blit_cctl_val = REG_FIELD_PREP(BLIT_CCTL_DST_MOCS_MASK, mocs_write_idx) |
>> - REG_FIELD_PREP(BLIT_CCTL_SRC_MOCS_MASK, mocs_read_idx);
>> struct xe_rtp_process_ctx ctx = XE_RTP_PROCESS_CTX_INITIALIZER(hwe);
>> - const struct xe_rtp_table_sr lrc_setup = XE_RTP_TABLE_SR(
>> - /*
>> - * Some blitter commands do not have a field for MOCS, those
>> - * commands will use MOCS index pointed by BLIT_CCTL.
>> - * BLIT_CCTL registers are needed to be programmed to un-cached.
>> - */
>> - { XE_RTP_NAME("BLIT_CCTL_default_MOCS"),
>> - XE_RTP_RULES(GRAPHICS_VERSION_RANGE(1200, 1274),
>> - ENGINE_CLASS(COPY)),
>> - XE_RTP_ACTIONS(FIELD_SET(BLIT_CCTL(0),
>> - BLIT_CCTL_DST_MOCS_MASK |
>> - BLIT_CCTL_SRC_MOCS_MASK,
>> - blit_cctl_val,
>> - XE_RTP_ACTION_FLAG(ENGINE_BASE)))
>> - },
>> - /* Disable WMTP if HW doesn't support it */
>> - { XE_RTP_NAME("DISABLE_WMTP_ON_UNSUPPORTED_HW"),
>> - XE_RTP_RULES(FUNC(xe_rtp_cfeg_wmtp_disabled)),
>> - XE_RTP_ACTIONS(FIELD_SET(CS_CHICKEN1(0),
>> - PREEMPT_GPGPU_LEVEL_MASK,
>> - PREEMPT_GPGPU_THREAD_GROUP_LEVEL)),
>> - XE_RTP_ENTRY_FLAG(FOREACH_ENGINE)
>> - },
>> - );
>>
>> xe_rtp_process_to_sr(&ctx, &lrc_setup, &hwe->reg_lrc, true);
>> }
>>
>> --
>> 2.54.0
>>
>
> --
> Matt Roper
> Graphics Software Engineer
> Linux GPU Platform Enablement
> Intel Corporation
next prev parent reply other threads:[~2026-06-16 22:27 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-16 19:21 [PATCH 0/4] Move RTP tables for engine init off the stack Matt Roper
2026-06-16 19:21 ` [PATCH 1/4] drm/xe: Reformat xe_rtp_types.h Matt Roper
2026-06-16 22:09 ` Gustavo Sousa
2026-06-16 19:21 ` [PATCH 2/4] drm/xe/rtp: Add FIELD_SET_FUNC RTP action Matt Roper
2026-06-16 22:22 ` Gustavo Sousa
2026-06-16 19:21 ` [PATCH 3/4] drm/xe: Move engines' LRC programming RTP table off the stack Matt Roper
2026-06-16 19:54 ` Matt Roper
2026-06-16 22:26 ` Gustavo Sousa [this message]
2026-06-16 19:21 ` [PATCH 4/4] drm/xe: Move engines' non-LRC " Matt Roper
2026-06-16 22:33 ` Gustavo Sousa
2026-06-16 19:28 ` ✗ CI.checkpatch: warning for Move RTP tables for engine init " Patchwork
2026-06-16 19:29 ` ✓ CI.KUnit: success " Patchwork
2026-06-16 20:19 ` ✓ Xe.CI.BAT: " Patchwork
2026-06-17 0:31 ` ✓ Xe.CI.FULL: " 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=87wlvyt93s.fsf@intel.com \
--to=gustavo.sousa@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=matthew.d.roper@intel.com \
--cc=violet.monti@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 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.