dma-buf: double unlock in debugfs code
Subsystems (2)
PATCH 1/2 dma-buf: double unlock in debugfs code
We unlock here when we failed to take the lock.
---
This is in linux-next, and I think the debugfs code is only in Sumit's tree.
drivers/base/dma-buf.c +1 -2
@@ -593,7 +593,7 @@ static int dma_buf_describe(struct seq_file *s)if (ret) {seq_printf (s,"\tERROR locking buffer object: skipping\n" ); - goto skip_buffer; +continue ; }seq_printf (s,"\t" );
@@ -618,7 +618,6 @@ static int dma_buf_describe(struct seq_file *s) count++; size += buf_obj->size; -skip_buffer:mutex_unlock (&buf_obj->lock); }
Oops... I mailed that prematurely. There isn't a [patch 2/2].
regards,
dan carpenter
Quoted 6 lines · Dan Carpenter
We unlock here when we failed to take the lock.
Signed-off-by Dan Carpenter <[email protected]>---
This is in linux-next, and I think the debugfs code is only in Sumit's tree.
Yes, it's not in mine, so I can't take this, sorry. Sumit, please apply this:
A message that is not in the archive (sent off-list or not yet received): CAO_48GHvAUmVmmZyHiEXzN3nWtTjC9iz-hEo-z3GrR2T-LZ-0w@mail.gmail.com
Quoted 9 lines · Sumit Semwal
Hi Dan,
Quoted 2 lines · Dan Carpenter
We unlock here when we failed to take the lock.
Thanks for catching this; I will add it to the for-next queue.
Could I merge this change with the main patch while submitting to mainline? With an attribution to you, of course.
Yeah. Just merge it. Don't stress about attribution.
regards,
dan carpenter