Messages in this thread | | | Date | Thu, 21 Mar 2013 10:18:38 -0400 | Subject | Re: [BUG] staging: android: ashmem: Deadlock during ashmem_mmap and ashmem_read | From | Robert Love <> |
| |
On Thu, Mar 21, 2013 at 10:06 AM, Bjorn Bringert <bringert@google.com> wrote: > I did implement ashmem_read, but I had no idea what I was doing. Calling the > VFS read function seemed like an obvious way to do it, but it might be > wrong. If that needs fixing, then the similar VFS call in ashmem_llseek > probably needs fixing too.
You don't want to hold ashmem_mutex across the VFS calls. It is only needed to protect the ashmem-internal structures.
FWIW is Android now using ashmem_read()? I left it out of the original ashmem implementation on purpose.
Robert
| |