dri-devel Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Qiang Yu <Qiang.Yu@amd.com>
To: amd-gfx@lists.freedesktop.org
Cc: Qiang Yu <Qiang.Yu@amd.com>,
	Emil Velikov <emil.l.velikov@gmail.com>,
	dri-devel@lists.freedesktop.org
Subject: [libdrm][PATCH v2 1/2] drm: Fix multi GPU drmGetDevice return wrong device
Date: Thu, 14 Jul 2016 17:10:55 +0800	[thread overview]
Message-ID: <1468487456-28666-1-git-send-email-Qiang.Yu@amd.com> (raw)

drmGetDevice will always return the first device it find
under /dev/dri/. This is not true for multi GPU situation.

Change-Id: I2a85a8a4feba8a5cc517ad75c6afb532fa07c53d
Signed-off-by: Qiang Yu <Qiang.Yu@amd.com>
---
 xf86drm.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/xf86drm.c b/xf86drm.c
index 6689f7c..19001db 100644
--- a/xf86drm.c
+++ b/xf86drm.c
@@ -3087,6 +3087,7 @@ int drmGetDevice(int fd, drmDevicePtr *device)
     int maj, min;
     int ret, i, node_count;
     int max_count = 16;
+    dev_t find_rdev;
 
     if (fd == -1 || device == NULL)
         return -EINVAL;
@@ -3094,6 +3095,7 @@ int drmGetDevice(int fd, drmDevicePtr *device)
     if (fstat(fd, &sbuf))
         return -errno;
 
+    find_rdev = sbuf.st_rdev;
     maj = major(sbuf.st_rdev);
     min = minor(sbuf.st_rdev);
 
@@ -3154,17 +3156,24 @@ int drmGetDevice(int fd, drmDevicePtr *device)
             local_devices = temp;
         }
 
-        local_devices[i] = d;
+        /* store target at local_devices[0] for ease to use below */
+        if (find_rdev == sbuf.st_rdev && i) {
+            local_devices[i] = local_devices[0];
+            local_devices[0] = d;
+        }
+        else
+            local_devices[i] = d;
         i++;
     }
     node_count = i;
 
-    /* Fold nodes into a single device if they share the same bus info */
+    /* Fold nodes into a single device if they share the same bus info
+     * and nodes with same bus info will be merged into the first node
+     * position in local_devices */
     drmFoldDuplicatedDevices(local_devices, node_count);
 
     *device = local_devices[0];
-    for (i = 1; i < node_count && local_devices[i]; i++)
-            drmFreeDevice(&local_devices[i]);
+    drmFreeDevices(&local_devices[1], node_count - 1);
 
     closedir(sysdir);
     free(local_devices);
-- 
1.9.1

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

             reply	other threads:[~2016-07-14  8:17 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-07-14  9:10 Qiang Yu [this message]
2016-07-14  9:10 ` [libdrm][PATCH v2 2/2] drm: fix multi GPU drmFreeDevices memory leak Qiang Yu
2016-07-14 17:20 ` [libdrm][PATCH v2 1/2] drm: Fix multi GPU drmGetDevice return wrong device Emil Velikov
2016-07-15  1:19   ` Yu, Qiang

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=1468487456-28666-1-git-send-email-Qiang.Yu@amd.com \
    --to=qiang.yu@amd.com \
    --cc=amd-gfx@lists.freedesktop.org \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=emil.l.velikov@gmail.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