From: Daniel Vetter <daniel@ffwll.ch>
To: oscar.mateo@intel.com
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH 13/53] drm/i915/bdw: Deferred creation of user-created LRCs
Date: Wed, 18 Jun 2014 22:27:30 +0200 [thread overview]
Message-ID: <20140618202730.GA5821@phenom.ffwll.local> (raw)
In-Reply-To: <1402673891-14618-14-git-send-email-oscar.mateo@intel.com>
On Fri, Jun 13, 2014 at 04:37:31PM +0100, oscar.mateo@intel.com wrote:
> From: Oscar Mateo <oscar.mateo@intel.com>
>
> The backing objects for contexts created via open fd are actually
> empty until the user starts sending execbuffers to them. We do this
> because, at create time, we really don't know which engine is going
> to be used with the context later on.
>
> v2: As context created via ioctl can only be used with the render ring,
> we have enough information to allocate & populate them right away.
Not sure this is a good choice to special-case the ioctl. We might want
to allow contexts also on non-render rings, and it introduces a special
case for not much gain. Since we must have the deferred alloca anyway lets
use that for all cases. And if there's a (performance) issue with it we
need to address that. One true (code) path to rule them all.
-Daniel
>
> Signed-off-by: Oscar Mateo <oscar.mateo@intel.com>
> ---
> drivers/gpu/drm/i915/i915_gem_context.c | 28 ++++++++++++++++++++++++++--
> 1 file changed, 26 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
> index 1fb4592..70bf6d0 100644
> --- a/drivers/gpu/drm/i915/i915_gem_context.c
> +++ b/drivers/gpu/drm/i915/i915_gem_context.c
> @@ -795,6 +795,7 @@ static bool contexts_enabled(struct drm_device *dev)
> int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
> struct drm_file *file)
> {
> + struct drm_i915_private *dev_priv = dev->dev_private;
> struct drm_i915_gem_context_create *args = data;
> struct drm_i915_file_private *file_priv = file->driver_priv;
> struct intel_context *ctx;
> @@ -808,9 +809,23 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
> return ret;
>
> ctx = i915_gem_create_context(dev, file_priv, USES_FULL_PPGTT(dev));
> - mutex_unlock(&dev->struct_mutex);
> - if (IS_ERR(ctx))
> + if (IS_ERR(ctx)) {
> + mutex_unlock(&dev->struct_mutex);
> return PTR_ERR(ctx);
> + }
> +
> + if (dev_priv->lrc_enabled) {
> + /* NB: We know this context will only be used with the render ring
> + * (as we enforce it) so we can allocate & populate it already */
> + int ret = intel_lr_context_deferred_create(ctx, &dev_priv->ring[RCS]);
> + if (ret) {
> + mutex_unlock(&dev->struct_mutex);
> + DRM_DEBUG_DRIVER("Could not create LRC: %d\n", ret);
> + return ret;
> + }
> + }
> +
> + mutex_unlock(&dev->struct_mutex);
>
> args->ctx_id = ctx->id;
> DRM_DEBUG_DRIVER("HW context %d created\n", args->ctx_id);
> @@ -851,6 +866,7 @@ struct intel_context *
> i915_gem_context_validate(struct drm_device *dev, struct drm_file *file,
> struct intel_engine_cs *ring, const u32 ctx_id)
> {
> + struct drm_i915_private *dev_priv = dev->dev_private;
> struct intel_context *ctx = NULL;
> struct i915_ctx_hang_stats *hs;
>
> @@ -867,5 +883,13 @@ i915_gem_context_validate(struct drm_device *dev, struct drm_file *file,
> return ERR_PTR(-EIO);
> }
>
> + if (dev_priv->lrc_enabled && !ctx->engine[ring->id].obj) {
> + int ret = intel_lr_context_deferred_create(ctx, ring);
> + if (ret) {
> + DRM_DEBUG("Could not create LRC %u: %d\n", ctx_id, ret);
> + return ERR_PTR(ret);
> + }
> + }
> +
> return ctx;
> }
> --
> 1.9.0
>
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
--
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
next prev parent reply other threads:[~2014-06-18 20:27 UTC|newest]
Thread overview: 149+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-06-13 15:37 [PATCH 00/53] Execlists v3 oscar.mateo
2014-06-13 15:37 ` [PATCH 01/53] drm/i915: Extract context backing object allocation oscar.mateo
2014-06-13 15:37 ` [PATCH 02/53] drm/i915: Rename ctx->obj to ctx->render_obj oscar.mateo
2014-06-13 17:00 ` Daniel Vetter
2014-06-16 15:20 ` Mateo Lozano, Oscar
2014-06-13 17:15 ` Chris Wilson
2014-06-13 15:37 ` [PATCH 03/53] drm/i915: Add a dev pointer to the context oscar.mateo
2014-06-13 15:37 ` [PATCH 04/53] drm/i915: Extract ringbuffer destroy & make alloc outside accesible oscar.mateo
2014-06-18 21:39 ` Volkin, Bradley D
2014-06-19 10:42 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 05/53] drm/i915: Move i915_gem_validate_context() to i915_gem_context.c oscar.mateo
2014-06-13 17:11 ` Chris Wilson
2014-06-16 15:18 ` Mateo Lozano, Oscar
2014-06-18 20:00 ` Volkin, Bradley D
2014-06-13 15:37 ` [PATCH 06/53] drm/i915/bdw: Introduce one context backing object per engine oscar.mateo
2014-06-18 20:16 ` Daniel Vetter
2014-06-19 8:52 ` Mateo Lozano, Oscar
2014-06-19 10:57 ` Daniel Vetter
2014-06-13 15:37 ` [PATCH 07/53] drm/i915/bdw: New file for Logical Ring Contexts and Execlists oscar.mateo
2014-06-18 20:17 ` Daniel Vetter
2014-06-19 9:01 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 08/53] drm/i915/bdw: Macro for LRCs and module option for Execlists oscar.mateo
2014-06-18 20:19 ` Daniel Vetter
2014-06-19 9:04 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 09/53] drm/i915/bdw: Initialization for Logical Ring Contexts oscar.mateo
2014-06-18 20:24 ` Daniel Vetter
2014-06-19 9:23 ` Mateo Lozano, Oscar
2014-06-19 10:08 ` Daniel Vetter
2014-06-19 10:10 ` Mateo Lozano, Oscar
2014-06-19 10:34 ` Daniel Vetter
2014-06-13 15:37 ` [PATCH 10/53] drm/i915/bdw: A bit more advanced context init/fini oscar.mateo
2014-06-18 22:13 ` Volkin, Bradley D
2014-06-19 6:13 ` Daniel Vetter
2014-06-13 15:37 ` [PATCH 11/53] drm/i915/bdw: Allocate ringbuffers for Logical Ring Contexts oscar.mateo
2014-06-18 22:19 ` Volkin, Bradley D
2014-06-23 12:07 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 12/53] drm/i915/bdw: Populate LR contexts (somewhat) oscar.mateo
2014-06-18 23:24 ` Volkin, Bradley D
2014-06-23 12:42 ` Mateo Lozano, Oscar
2014-06-23 15:05 ` Volkin, Bradley D
2014-06-23 15:11 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 13/53] drm/i915/bdw: Deferred creation of user-created LRCs oscar.mateo
2014-06-18 20:27 ` Daniel Vetter [this message]
2014-06-13 15:37 ` [PATCH 14/53] drm/i915/bdw: Render moot context reset and switch when LRCs are enabled oscar.mateo
2014-06-13 15:37 ` [PATCH 15/53] drm/i915/bdw: Don't write PDP in the legacy way when using LRCs oscar.mateo
2014-06-18 23:42 ` Volkin, Bradley D
2014-06-23 12:45 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 16/53] drm/i915/bdw: Skeleton for the new logical rings submission path oscar.mateo
2014-06-13 15:37 ` [PATCH 17/53] drm/i915/bdw: Generic logical ring init and cleanup oscar.mateo
2014-06-13 15:37 ` [PATCH 18/53] drm/i915/bdw: New header file for LRs, LRCs and Execlists oscar.mateo
2014-06-13 15:37 ` [PATCH 19/53] drm/i915: Extract pipe control fini & make init outside accesible oscar.mateo
2014-06-18 20:31 ` Daniel Vetter
2014-06-19 0:04 ` Volkin, Bradley D
2014-06-19 10:58 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 20/53] drm/i915/bdw: GEN-specific logical ring init oscar.mateo
2014-06-13 15:37 ` [PATCH 21/53] drm/i915/bdw: GEN-specific logical ring set/get seqno oscar.mateo
2014-06-13 15:37 ` [PATCH 22/53] drm/i915: Make ring_space more generic and outside accesible oscar.mateo
2014-06-13 15:37 ` [PATCH 23/53] drm/i915: Generalize intel_ring_get_tail oscar.mateo
2014-06-20 20:17 ` Volkin, Bradley D
2014-06-13 15:37 ` [PATCH 24/53] drm/i915: Make intel_ring_stopped outside accesible oscar.mateo
2014-06-13 15:37 ` [PATCH 25/53] drm/i915/bdw: GEN-specific logical ring submit context (somewhat) oscar.mateo
2014-06-20 20:28 ` Volkin, Bradley D
2014-06-23 12:49 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 26/53] drm/i915/bdw: New logical ring submission mechanism oscar.mateo
2014-06-20 21:00 ` Volkin, Bradley D
2014-06-23 13:09 ` Mateo Lozano, Oscar
2014-06-23 13:13 ` Chris Wilson
2014-06-23 13:18 ` Mateo Lozano, Oscar
2014-06-23 13:27 ` Chris Wilson
2014-06-23 13:36 ` Mateo Lozano, Oscar
2014-06-23 13:41 ` Chris Wilson
2014-06-23 14:35 ` Mateo Lozano, Oscar
2014-06-23 19:10 ` Volkin, Bradley D
2014-06-24 12:29 ` Mateo Lozano, Oscar
2014-07-07 12:39 ` Daniel Vetter
2014-06-24 0:23 ` Ben Widawsky
2014-06-24 11:45 ` Mateo Lozano, Oscar
2014-06-24 14:41 ` Volkin, Bradley D
2014-06-24 17:19 ` Jesse Barnes
2014-06-26 13:28 ` Mateo Lozano, Oscar
2014-07-07 12:41 ` Daniel Vetter
2014-06-13 15:37 ` [PATCH 27/53] drm/i915/bdw: GEN-specific logical ring emit request oscar.mateo
2014-06-20 21:18 ` Volkin, Bradley D
2014-06-23 15:48 ` Mateo Lozano, Oscar
2014-06-13 15:37 ` [PATCH 28/53] drm/i915/bdw: GEN-specific logical ring emit flush oscar.mateo
2014-06-20 21:39 ` Volkin, Bradley D
2014-06-13 15:37 ` [PATCH 29/53] drm/i915/bdw: Emission of requests with logical rings oscar.mateo
2014-06-13 15:37 ` [PATCH 30/53] drm/i915/bdw: Ring idle and stop " oscar.mateo
2014-06-13 15:37 ` [PATCH 31/53] drm/i915/bdw: Interrupts " oscar.mateo
2014-06-13 15:37 ` [PATCH 32/53] drm/i915/bdw: GEN-specific logical ring emit batchbuffer start oscar.mateo
2014-06-13 15:37 ` [PATCH 33/53] drm/i915: Extract the actual workload submission mechanism from execbuffer oscar.mateo
2014-06-13 15:37 ` [PATCH 34/53] drm/i915: Make move_to_active and retire_commands outside accesible oscar.mateo
2014-06-13 15:37 ` [PATCH 35/53] drm/i915/bdw: Workload submission mechanism for Execlists oscar.mateo
2014-06-13 15:37 ` [PATCH 36/53] drm/i915: Abstract the workload submission mechanism away oscar.mateo
2014-06-18 20:40 ` Daniel Vetter
2014-06-13 15:37 ` [PATCH 37/53] drm/i915/bdw: Implement context switching (somewhat) oscar.mateo
2014-06-13 17:00 ` Chris Wilson
2014-06-13 15:37 ` [PATCH 38/53] drm/i915/bdw: Write the tail pointer, LRC style oscar.mateo
2014-06-13 15:37 ` [PATCH 39/53] drm/i915/bdw: Two-stage execlist submit process oscar.mateo
2014-06-13 15:37 ` [PATCH 40/53] drm/i915/bdw: Handle context switch events oscar.mateo
2014-06-13 15:37 ` [PATCH 41/53] drm/i915/bdw: Avoid non-lite-restore preemptions oscar.mateo
2014-06-18 20:49 ` Daniel Vetter
2014-06-23 11:52 ` Mateo Lozano, Oscar
2014-07-07 12:47 ` Daniel Vetter
2014-06-13 15:38 ` [PATCH 42/53] drm/i915/bdw: Make sure gpu reset still works with Execlists oscar.mateo
2014-06-18 20:50 ` Daniel Vetter
2014-06-19 9:37 ` Mateo Lozano, Oscar
2014-06-13 15:38 ` [PATCH 43/53] drm/i915/bdw: Make sure error capture keeps working " oscar.mateo
2014-06-13 16:54 ` Chris Wilson
2014-06-18 20:52 ` Daniel Vetter
2014-06-18 20:53 ` Daniel Vetter
2014-06-13 15:38 ` [PATCH 44/53] drm/i915/bdw: Help out the ctx switch interrupt handler oscar.mateo
2014-06-13 15:38 ` [PATCH 45/53] drm/i915/bdw: Do not call intel_runtime_pm_get() in an interrupt oscar.mateo
2014-06-18 20:54 ` Daniel Vetter
2014-07-26 10:27 ` Chris Wilson
2014-07-28 8:54 ` Daniel Vetter
2014-07-29 7:37 ` Chris Wilson
2014-07-29 10:26 ` Daniel Vetter
2014-08-08 9:20 ` Chris Wilson
2014-08-08 9:37 ` Daniel Vetter
2014-08-08 13:41 ` Greg KH
2014-08-09 0:18 ` Rafael J. Wysocki
2014-08-09 0:14 ` Rafael J. Wysocki
2014-08-09 1:21 ` [Intel-gfx] " Alan Stern
2014-08-09 8:53 ` Daniel Vetter
2014-08-10 1:55 ` Rafael J. Wysocki
2014-06-13 15:38 ` [PATCH 46/53] drm/i915/bdw: Display execlists info in debugfs oscar.mateo
2014-06-18 20:59 ` Daniel Vetter
2014-06-13 15:38 ` [PATCH 47/53] drm/i915/bdw: Display context backing obj & ringbuffer " oscar.mateo
2014-06-13 15:38 ` [PATCH 48/53] drm/i915/bdw: Print context state " oscar.mateo
2014-06-13 15:38 ` [PATCH 49/53] drm/i915: Extract render state preparation oscar.mateo
2014-06-13 15:38 ` [PATCH 50/53] drm/i915/bdw: Render state init for Execlists oscar.mateo
2014-06-13 15:38 ` [PATCH 51/53] drm/i915/bdw: Document Logical Rings, LR contexts and Execlists oscar.mateo
2014-06-13 16:51 ` Chris Wilson
2014-06-16 15:24 ` Mateo Lozano, Oscar
2014-06-16 17:56 ` Daniel Vetter
2014-06-17 8:22 ` Mateo Lozano, Oscar
2014-06-17 9:39 ` Daniel Vetter
2014-06-17 9:46 ` Mateo Lozano, Oscar
2014-06-17 10:08 ` Daniel Vetter
2014-06-17 10:12 ` Mateo Lozano, Oscar
2014-06-13 15:38 ` [PATCH 52/53] drm/i915/bdw: Enable logical ring contexts oscar.mateo
2014-06-13 15:38 ` [PATCH 53/53] !UPSTREAM: drm/i915: Use MMIO flips oscar.mateo
2014-06-18 21:01 ` Daniel Vetter
2014-06-19 9:50 ` Mateo Lozano, Oscar
2014-06-19 10:04 ` Daniel Vetter
2014-06-19 10:13 ` Chris Wilson
2014-06-19 10:33 ` Mateo Lozano, Oscar
2014-06-18 21:26 ` [PATCH 00/53] Execlists v3 Daniel Vetter
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=20140618202730.GA5821@phenom.ffwll.local \
--to=daniel@ffwll.ch \
--cc=intel-gfx@lists.freedesktop.org \
--cc=oscar.mateo@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