From: "Christian König" <ckoenig.leichtzumerken@gmail.com>
To: dmoulding@me.com, sf@sfritsch.de, bskeggs@redhat.com
Cc: nouveau@lists.freedesktop.org, dri-devel@lists.freedesktop.org
Subject: [PATCH] drm/nouveau: wait for the exclusive fence after the shared ones v2
Date: Thu, 9 Dec 2021 11:23:35 +0100 [thread overview]
Message-ID: <20211209102335.18321-1-christian.koenig@amd.com> (raw)
Always waiting for the exclusive fence resulted on some performance
regressions. So try to wait for the shared fences first, then the
exclusive fence should always be signaled already.
v2: fix incorrectly placed "(", add some comment why we do this.
Signed-off-by: Christian König <christian.koenig@amd.com>
---
drivers/gpu/drm/nouveau/nouveau_fence.c | 28 +++++++++++++------------
1 file changed, 15 insertions(+), 13 deletions(-)
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 05d0b3eb3690..0ae416aa76dc 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -353,15 +353,22 @@ nouveau_fence_sync(struct nouveau_bo *nvbo, struct nouveau_channel *chan, bool e
if (ret)
return ret;
- }
- fobj = dma_resv_shared_list(resv);
- fence = dma_resv_excl_fence(resv);
+ fobj = NULL;
+ } else {
+ fobj = dma_resv_shared_list(resv);
+ }
- if (fence) {
+ /* Waiting for the exclusive fence first causes performance regressions
+ * under some circumstances. So manually wait for the shared ones first.
+ */
+ for (i = 0; i < (fobj ? fobj->shared_count : 0) && !ret; ++i) {
struct nouveau_channel *prev = NULL;
bool must_wait = true;
+ fence = rcu_dereference_protected(fobj->shared[i],
+ dma_resv_held(resv));
+
f = nouveau_local_fence(fence, chan->drm);
if (f) {
rcu_read_lock();
@@ -373,20 +380,13 @@ nouveau_fence_sync(struct nouveau_bo *nvbo, struct nouveau_channel *chan, bool e
if (must_wait)
ret = dma_fence_wait(fence, intr);
-
- return ret;
}
- if (!exclusive || !fobj)
- return ret;
-
- for (i = 0; i < fobj->shared_count && !ret; ++i) {
+ fence = dma_resv_excl_fence(resv);
+ if (fence) {
struct nouveau_channel *prev = NULL;
bool must_wait = true;
- fence = rcu_dereference_protected(fobj->shared[i],
- dma_resv_held(resv));
-
f = nouveau_local_fence(fence, chan->drm);
if (f) {
rcu_read_lock();
@@ -398,6 +398,8 @@ nouveau_fence_sync(struct nouveau_bo *nvbo, struct nouveau_channel *chan, bool e
if (must_wait)
ret = dma_fence_wait(fence, intr);
+
+ return ret;
}
return ret;
--
2.25.1
next reply other threads:[~2021-12-09 16:57 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-12-09 10:23 Christian König [this message]
2021-12-10 9:06 ` [Nouveau] [PATCH] drm/nouveau: wait for the exclusive fence after the shared ones v2 Thorsten Leemhuis
2021-12-11 9:59 ` Stefan Fritsch
2021-12-14 9:19 ` Christian König
2021-12-15 22:32 ` [Nouveau] " Ben Skeggs
2021-12-21 10:11 ` Thorsten Leemhuis
2021-12-21 10:20 ` Christian König
2021-12-20 19:17 ` Dan Moulding
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=20211209102335.18321-1-christian.koenig@amd.com \
--to=ckoenig.leichtzumerken@gmail.com \
--cc=bskeggs@redhat.com \
--cc=dmoulding@me.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=nouveau@lists.freedesktop.org \
--cc=sf@sfritsch.de \
/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).