From: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>, intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 09/33] drm/i915: Mark unmappable GGTT entries as PIN_HIGH
Date: Tue, 9 Aug 2016 12:05:30 +0100 [thread overview]
Message-ID: <57A9B8FA.2080504@linux.intel.com> (raw)
In-Reply-To: <1470581141-14432-10-git-send-email-chris@chris-wilson.co.uk>
On 07/08/16 15:45, Chris Wilson wrote:
> We allocate a few objects into the GGTT that we never need to access via
> the mappable aperture (such as contexts, status pages). We can request
> that these are bound high in the VM to increase the amount of mappable
> aperture available. However, anything that may be frequently pinned
> (such as logical contexts) we want to use the fast search & insert.
Is the last bit still true after you merged:
commit 202b52b7fbf70858609ec20829c7d69a13ffa351
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Wed Aug 3 16:04:09 2016 +0100
drm: Track drm_mm nodes with an interval tree
Or we could PIN_HIGH the LRCs as well now?
Regards,
Tvrtko
> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
> ---
> drivers/gpu/drm/i915/intel_lrc.c | 2 +-
> drivers/gpu/drm/i915/intel_ringbuffer.c | 5 +++--
> 2 files changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
> index 309c5d9b1c57..c7f4b64b16f6 100644
> --- a/drivers/gpu/drm/i915/intel_lrc.c
> +++ b/drivers/gpu/drm/i915/intel_lrc.c
> @@ -1182,7 +1182,7 @@ static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size)
> }
>
> ret = i915_gem_object_ggtt_pin(engine->wa_ctx.obj, NULL,
> - 0, PAGE_SIZE, 0);
> + 0, PAGE_SIZE, PIN_HIGH);
> if (ret) {
> DRM_DEBUG_DRIVER("pin LRC WA ctx backing obj failed: %d\n",
> ret);
> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
> index 16b726fe33eb..09f01c641c14 100644
> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
> @@ -2093,7 +2093,7 @@ static int intel_ring_context_pin(struct i915_gem_context *ctx,
>
> if (ce->state) {
> ret = i915_gem_object_ggtt_pin(ce->state, NULL, 0,
> - ctx->ggtt_alignment, 0);
> + ctx->ggtt_alignment, PIN_HIGH);
> if (ret)
> goto error;
> }
> @@ -2629,7 +2629,8 @@ static void intel_ring_init_semaphores(struct drm_i915_private *dev_priv,
> i915.semaphores = 0;
> } else {
> i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
> - ret = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, 0);
> + ret = i915_gem_object_ggtt_pin(obj, NULL,
> + 0, 0, PIN_HIGH);
> if (ret != 0) {
> i915_gem_object_put(obj);
> DRM_ERROR("Failed to pin semaphore bo. Disabling semaphores\n");
>
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
next prev parent reply other threads:[~2016-08-09 11:05 UTC|newest]
Thread overview: 125+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-08-07 14:45 First class VMA, take 2 Chris Wilson
2016-08-07 14:45 ` [PATCH 01/33] drm/i915: Add smp_rmb() to busy ioctl's RCU dance Chris Wilson
2016-08-08 9:12 ` Daniel Vetter
2016-08-08 9:30 ` Chris Wilson
2016-08-08 9:45 ` Chris Wilson
2016-08-09 6:36 ` Joonas Lahtinen
2016-08-09 7:14 ` Chris Wilson
2016-08-09 8:48 ` Joonas Lahtinen
2016-08-09 9:05 ` Chris Wilson
2016-08-10 10:12 ` Daniel Vetter
2016-08-10 10:13 ` Daniel Vetter
2016-08-10 11:00 ` Joonas Lahtinen
2016-08-12 9:50 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 02/33] drm/i915: Do not overwrite the request with zero on reallocation Chris Wilson
2016-08-08 9:25 ` Daniel Vetter
2016-08-08 9:56 ` Chris Wilson
2016-08-09 6:32 ` Daniel Vetter
2016-08-07 14:45 ` [PATCH 03/33] drm/i915: Move missed interrupt detection from hangcheck to breadcrumbs Chris Wilson
2016-08-09 14:08 ` [PATCH v2] " Chris Wilson
2016-08-09 14:10 ` [PATCH v3] " Chris Wilson
2016-08-09 15:24 ` Mika Kuoppala
2016-08-07 14:45 ` [PATCH 04/33] drm/i915: Use RCU to annotate and enforce protection for breadcrumb's bh Chris Wilson
2016-08-08 9:33 ` Daniel Vetter
2016-08-12 9:56 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 05/33] drm/i915: Reduce amount of duplicate buffer information captured on error Chris Wilson
2016-08-10 7:04 ` Joonas Lahtinen
2016-08-10 7:15 ` Chris Wilson
2016-08-10 8:07 ` Joonas Lahtinen
2016-08-10 8:36 ` Chris Wilson
2016-08-10 10:51 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 06/33] drm/i915: Stop the machine whilst capturing the GPU crash dump Chris Wilson
2016-08-07 14:45 ` [PATCH 07/33] drm/i915: Store the active context object on all engines upon error Chris Wilson
2016-08-09 9:02 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 08/33] drm/i915: Move setting of request->batch into its single callsite Chris Wilson
2016-08-09 15:53 ` Mika Kuoppala
2016-08-09 16:04 ` Chris Wilson
2016-08-10 7:19 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 09/33] drm/i915: Mark unmappable GGTT entries as PIN_HIGH Chris Wilson
2016-08-08 9:09 ` Joonas Lahtinen
2016-08-09 11:05 ` Tvrtko Ursulin [this message]
2016-08-09 11:13 ` Chris Wilson
2016-08-09 11:20 ` Chris Wilson
2016-08-07 14:45 ` [PATCH 10/33] drm/i915: Remove inactive/active list from debugfs Chris Wilson
2016-08-09 10:29 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 11/33] drm/i915: Focus debugfs/i915_gem_pinned to show only display pins Chris Wilson
2016-08-09 10:39 ` Joonas Lahtinen
2016-08-09 10:46 ` Chris Wilson
2016-08-09 11:32 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 12/33] drm/i915: Reduce i915_gem_objects to only show object information Chris Wilson
2016-08-10 7:29 ` Joonas Lahtinen
2016-08-10 7:38 ` Chris Wilson
2016-08-10 8:10 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 13/33] drm/i915: Remove redundant WARN_ON from __i915_add_request() Chris Wilson
2016-08-08 9:03 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 14/33] drm/i915: Create a VMA for an object Chris Wilson
2016-08-08 9:01 ` Joonas Lahtinen
2016-08-08 9:09 ` Chris Wilson
2016-08-10 10:58 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 15/33] drm/i915: Track pinned vma inside guc Chris Wilson
2016-08-11 16:19 ` Dave Gordon
2016-08-11 16:41 ` Chris Wilson
2016-08-07 14:45 ` [PATCH 16/33] drm/i915: Convert fence computations to use vma directly Chris Wilson
2016-08-09 10:27 ` Joonas Lahtinen
2016-08-09 10:33 ` Chris Wilson
2016-08-07 14:45 ` [PATCH 17/33] drm/i915: Use VMA directly for checking tiling parameters Chris Wilson
2016-08-09 6:18 ` Joonas Lahtinen
2016-08-09 8:03 ` Chris Wilson
2016-08-07 14:45 ` [PATCH 18/33] drm/i915: Use VMA as the primary object for context state Chris Wilson
2016-08-10 8:03 ` Joonas Lahtinen
2016-08-10 8:25 ` Chris Wilson
2016-08-10 10:54 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 19/33] drm/i915: Only clflush the context object when binding Chris Wilson
2016-08-10 8:41 ` Joonas Lahtinen
2016-08-10 9:02 ` Chris Wilson
2016-08-10 10:50 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 20/33] drm/i915: Use VMA for ringbuffer tracking Chris Wilson
2016-08-11 9:32 ` Joonas Lahtinen
2016-08-11 9:58 ` Chris Wilson
2016-08-07 14:45 ` [PATCH 21/33] drm/i915: Use VMA for scratch page tracking Chris Wilson
2016-08-08 8:00 ` [PATCH 1/3] " Chris Wilson
2016-08-08 8:00 ` [PATCH 2/3] drm/i915: Move common scratch allocation/destroy to intel_engine_cs.c Chris Wilson
2016-08-08 9:24 ` Matthew Auld
2016-08-08 8:00 ` [PATCH 3/3] drm/i915: Move common seqno reset " Chris Wilson
2016-08-08 9:40 ` Matthew Auld
2016-08-08 10:15 ` Chris Wilson
2016-08-08 15:34 ` Matthew Auld
2016-08-11 10:06 ` [PATCH 21/33] drm/i915: Use VMA for scratch page tracking Joonas Lahtinen
2016-08-11 10:22 ` Chris Wilson
2016-08-07 14:45 ` [PATCH 22/33] drm/i915/overlay: Use VMA as the primary tracker for images Chris Wilson
2016-08-11 10:17 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 23/33] drm/i915: Use VMA as the primary tracker for semaphore page Chris Wilson
2016-08-11 10:42 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 24/33] drm/i915: Use VMA for render state page tracking Chris Wilson
2016-08-11 10:46 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 25/33] drm/i915: Use VMA for wa_ctx tracking Chris Wilson
2016-08-11 10:53 ` Joonas Lahtinen
2016-08-11 11:02 ` Chris Wilson
2016-08-11 12:41 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 26/33] drm/i915: Track pinned VMA Chris Wilson
2016-08-11 12:18 ` Joonas Lahtinen
2016-08-11 12:37 ` Chris Wilson
2016-08-07 14:45 ` [PATCH 27/33] drm/i915: Print the batchbuffer offset next to BBADDR in error state Chris Wilson
2016-08-11 12:24 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 28/33] drm/i915: Move per-request pid from request to ctx Chris Wilson
2016-08-11 12:32 ` Joonas Lahtinen
2016-08-11 12:41 ` Chris Wilson
2016-08-07 14:45 ` [PATCH 29/33] drm/i915: Only record active and pending requests upon a GPU hang Chris Wilson
2016-08-11 12:36 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 30/33] drm/i915: Record the RING_MODE register for post-mortem debugging Chris Wilson
2016-08-08 11:35 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 31/33] drm/i915: Always use the GTT for error capture Chris Wilson
2016-08-07 14:45 ` [PATCH 32/33] drm/i915: Consolidate error object printing Chris Wilson
2016-08-09 11:44 ` Joonas Lahtinen
2016-08-09 11:53 ` Chris Wilson
2016-08-10 10:55 ` Joonas Lahtinen
2016-08-07 14:45 ` [PATCH 33/33] drm/i915: Compress GPU objects in error state Chris Wilson
2016-08-10 10:32 ` Joonas Lahtinen
2016-08-10 10:52 ` Chris Wilson
2016-08-10 11:26 ` Joonas Lahtinen
2016-08-07 15:16 ` ✗ Ro.CI.BAT: failure for series starting with [01/33] drm/i915: Add smp_rmb() to busy ioctl's RCU dance Patchwork
2016-08-08 9:46 ` ✗ Ro.CI.BAT: failure for series starting with [01/33] drm/i915: Add smp_rmb() to busy ioctl's RCU dance (rev4) Patchwork
2016-08-08 10:34 ` ✗ Fi.CI.BAT: " Patchwork
2016-08-09 14:10 ` ✗ Ro.CI.BAT: failure for series starting with [01/33] drm/i915: Add smp_rmb() to busy ioctl's RCU dance (rev5) Patchwork
2016-08-09 14:20 ` ✗ Ro.CI.BAT: failure for series starting with [01/33] drm/i915: Add smp_rmb() to busy ioctl's RCU dance (rev6) Patchwork
2016-08-10 6:43 ` 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=57A9B8FA.2080504@linux.intel.com \
--to=tvrtko.ursulin@linux.intel.com \
--cc=chris@chris-wilson.co.uk \
--cc=intel-gfx@lists.freedesktop.org \
/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