Message ID | 20190927102318.12830-1-nborisov@suse.com (mailing list archive) |
---|---|
Headers | show |
Series | btrfs llseek improvement, take 2 | expand |
On Fri, Sep 27, 2019 at 01:23:15PM +0300, Nikolay Borisov wrote: > Here is v2 of the llseek improvements, main changes are: > > * Patch 1 - changed the locking scheme. I'm now using inode_lock_shared since > holding the extent lock is not sufficient to prevent concurrent truncates. Regarding the offline discussions, I'd like to see more people look at that until we're sure that it works, namely seek vs truncate and extent range locking.
On Fri, Sep 27, 2019 at 01:23:15PM +0300, Nikolay Borisov wrote:
> Here is v2 of the llseek improvements, main changes are:
Btw, with Goldwyn looking into btrfs iomap support wouldn't it make
sense to try to reuse the iomap lseek code?
On 27.09.19 г. 20:03 ч., Christoph Hellwig wrote: >> Here is v2 of the llseek improvements, main changes are: > Btw, with Goldwyn looking into btrfs iomap support wouldn't it make > sense to try to reuse the iomap lseek code? > When that code lands - yes. ATM - there are more pressing stuff.
On Fri, Sep 27, 2019 at 01:23:15PM +0300, Nikolay Borisov wrote: > Here is v2 of the llseek improvements, main changes are: > > * Patch 1 - changed the locking scheme. I'm now using inode_lock_shared since > holding the extent lock is not sufficient to prevent concurrent truncates. > > * Fixed lingo bugs in patch 2 changelog (Johaness) > > Nikolay Borisov (3): > btrfs: Speed up btrfs_file_llseek > btrfs: Simplify btrfs_file_llseek > btrfs: Return offset from find_desired_extent Moved from topic branch to misc-next. Thanks.