intel-gfx.lists.freedesktop.org archive mirror
 help / color / mirror / Atom feed
From: Dave Gordon <david.s.gordon@intel.com>
To: intel-gfx@lists.freedesktop.org
Subject: [PATCH v2 5/6] drm/i915: HWSP should be unmapped earlier in LRC teardown sequence
Date: Fri, 22 Jan 2016 23:10:10 +0000	[thread overview]
Message-ID: <1453504211-7982-6-git-send-email-david.s.gordon@intel.com> (raw)
In-Reply-To: <1453504211-7982-1-git-send-email-david.s.gordon@intel.com>

In LRC mode, the HWSP is part of the default context object, and
therefore does not exist independently. Worse, it doesn't contribute
to the refcount on the default context object either.

Currently, the default context is deallocated in intel_lr_context_free(),
but the HWSP kmapping is not torn down until the subsequent call to
intel_logical_ring_cleanup(). Between these calls, ring->status_page.obj
continues to point to the (now non-existent) default context object,
and the kernel mapping likewise points to a page which is now free.

The solution is to dispose of the HWSP kmapping and pointer before the
object itself is freed, so this patch moves the kmap teardown code from
intel_lr_context_free() to intel_logical_ring_cleanup().

This code was introduced in

    48d8238 drm/i915/bdw: Generic logical ring init and cleanup

i.e. it has been there ever since LRC mode was first implemented.

v2:
    Split from "handle teardown of HWSP when context is freed".

Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
---
 drivers/gpu/drm/i915/intel_lrc.c | 26 +++++++++++++++++++-------
 1 file changed, 19 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index a3ab2b4..c702fc2 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1986,13 +1986,9 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *ring)
 	i915_cmd_parser_fini_ring(ring);
 	i915_gem_batch_pool_fini(&ring->batch_pool);
 
-	if (ring->status_page.obj) {
-		struct page *page;
-
-		page = i915_gem_object_get_page(ring->status_page.obj, LRC_PPHWSP_PN);
-		kunmap(page);
-		ring->status_page.obj = NULL;
-	}
+	/* Status page should have gone already */
+	WARN_ON(ring->status_page.page_addr);
+	WARN_ON(ring->status_page.obj);
 
 	ring->disable_lite_restore_wa = false;
 	ring->ctx_desc_template = 0;
@@ -2431,6 +2427,22 @@ void intel_lr_context_free(struct intel_context *ctx)
 			continue;
 
 		if (ctx == ctx->i915->kernel_context) {
+			struct intel_engine_cs *ring = ringbuf->ring;
+
+			/*
+			 * The HWSP is part of the kernel context
+			 * object in LRC mode, so tear it down now.
+			 */
+			if (ring->status_page.obj) {
+				struct page *page;
+
+				page = i915_gem_object_get_page(
+						ring->status_page.obj,
+						LRC_PPHWSP_PN);
+				kunmap(page);
+				ring->status_page.obj = NULL;
+			}
+
 			intel_unpin_ringbuffer_obj(ringbuf);
 			i915_gem_object_ggtt_unpin(ctx_obj);
 		}
-- 
1.9.1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

  parent reply	other threads:[~2016-01-22 23:10 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-01-22 23:10 [PATCH v2 0/6] A collection of cleanups Dave Gordon
2016-01-22 23:10 ` [PATCH v2 1/6] drm/i915: fix context/engine cleanup order Dave Gordon
2016-02-10  7:40   ` Daniel Vetter
2016-01-22 23:10 ` [PATCH v2 2/6] drm/i915: tidy up initialisation failure paths (GEM & LRC) Dave Gordon
2016-01-22 23:10 ` [PATCH v2 3/6] drm/i915: tidy up initialisation failure paths (legacy) Dave Gordon
2016-01-25 10:52   ` Chris Wilson
2016-01-25 12:08     ` Dave Gordon
2016-01-25 18:23       ` Daniel Vetter
2016-01-22 23:10 ` [PATCH v2 4/6] drm/i915: unmap the correct page in intel_logical_ring_cleanup() Dave Gordon
2016-01-25 10:54   ` Chris Wilson
2016-01-25 12:15     ` Dave Gordon
2016-01-22 23:10 ` Dave Gordon [this message]
2016-01-25 10:56   ` [PATCH v2 5/6] drm/i915: HWSP should be unmapped earlier in LRC teardown sequence Chris Wilson
2016-01-25 12:27     ` Dave Gordon
2016-01-22 23:10 ` [PATCH v2 6/6] drm/i915: consolidate LRC mode HWSP setup & teardown Dave Gordon
2016-01-23  7:43 ` ✓ Fi.CI.BAT: success for A collection of cleanups Patchwork
2016-01-28 12:13 ` 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=1453504211-7982-6-git-send-email-david.s.gordon@intel.com \
    --to=david.s.gordon@intel.com \
    --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;
as well as URLs for NNTP newsgroup(s).