[PATCH 2/4] drm/xe/ct: prevent UAF in send_recv()
Timo Aaltonen
tjaalton at ubuntu.com
Wed Oct 30 14:51:12 UTC 2024
From: Matthew Auld <matthew.auld at intel.com>
BugLink: https://bugs.launchpad.net/bugs/2085944
Ensure we serialize with completion side to prevent UAF with fence going
out of scope on the stack, since we have no clue if it will fire after
the timeout before we can erase from the xa. Also we have some dependent
loads and stores for which we need the correct ordering, and we lack the
needed barriers. Fix this by grabbing the ct->lock after the wait, which
is also held by the completion side.
v2 (Badal):
- Also print done after acquiring the lock and seeing timeout.
Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
Signed-off-by: Matthew Auld <matthew.auld at intel.com>
Cc: Matthew Brost <matthew.brost at intel.com>
Cc: Badal Nilawar <badal.nilawar at intel.com>
Cc: <stable at vger.kernel.org> # v6.8+
Reviewed-by: Badal Nilawar <badal.nilawar at intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241001084346.98516-5-matthew.auld@intel.com
(cherry picked from commit 52789ce35c55ccd30c4b67b9cc5b2af55e0122ea)
Signed-off-by: Timo Aaltonen <timo.aaltonen at canonical.com>
---
drivers/gpu/drm/xe/xe_guc_ct.c | 21 ++++++++++++++++++---
1 file changed, 18 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/xe/xe_guc_ct.c b/drivers/gpu/drm/xe/xe_guc_ct.c
index 64afc90ad2c5..ef5059db8aaa 100644
--- a/drivers/gpu/drm/xe/xe_guc_ct.c
+++ b/drivers/gpu/drm/xe/xe_guc_ct.c
@@ -894,16 +894,26 @@ static int guc_ct_send_recv(struct xe_guc_ct *ct, const u32 *action, u32 len,
}
ret = wait_event_timeout(ct->g2h_fence_wq, g2h_fence.done, HZ);
+
+ /*
+ * Ensure we serialize with completion side to prevent UAF with fence going out of scope on
+ * the stack, since we have no clue if it will fire after the timeout before we can erase
+ * from the xa. Also we have some dependent loads and stores below for which we need the
+ * correct ordering, and we lack the needed barriers.
+ */
+ mutex_lock(&ct->lock);
if (!ret) {
- xe_gt_err(gt, "Timed out wait for G2H, fence %u, action %04x",
- g2h_fence.seqno, action[0]);
+ xe_gt_err(gt, "Timed out wait for G2H, fence %u, action %04x, done %s",
+ g2h_fence.seqno, action[0], str_yes_no(g2h_fence.done));
xa_erase_irq(&ct->fence_lookup, g2h_fence.seqno);
+ mutex_unlock(&ct->lock);
return -ETIME;
}
if (g2h_fence.retry) {
xe_gt_dbg(gt, "H2G action %#x retrying: reason %#x\n",
action[0], g2h_fence.reason);
+ mutex_unlock(&ct->lock);
goto retry;
}
if (g2h_fence.fail) {
@@ -912,7 +922,12 @@ static int guc_ct_send_recv(struct xe_guc_ct *ct, const u32 *action, u32 len,
ret = -EIO;
}
- return ret > 0 ? response_buffer ? g2h_fence.response_len : g2h_fence.response_data : ret;
+ if (ret > 0)
+ ret = response_buffer ? g2h_fence.response_len : g2h_fence.response_data;
+
+ mutex_unlock(&ct->lock);
+
+ return ret;
}
/**
--
2.45.2
More information about the kernel-team
mailing list