Messages in this thread | | | Date | Thu, 3 Jun 2010 03:06:08 -0400 | From | Christoph Hellwig <> | Subject | Re: [RFC 4/5] BKL: use no BKL in llseek |
| |
On Thu, Jun 03, 2010 at 02:13:18AM +0200, Arnd Bergmann wrote: > We have shown that the BKL in default_llseek and other > llseek operations never protects against concurrent access > from another function:
From VFS POV no to the default_llseek conversion. As told you and others of the BKL brigade don't change this thing but get rid of it entirely. I was under the impression we had agreed on a road map for that anyway. Please don't forget that overall kernel improvement should at least be a side effect of your big sweap and clean.
| |