[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: [VOTE] Merging 1.9-cache-improvements to /trunk

From: Stefan Fuhrmann <stefan.fuhrmann_at_wandisco.com>
Date: Sun, 17 May 2015 16:45:31 +0200

On Sat, May 16, 2015 at 7:45 PM, Ivan Zhakov <ivan_at_visualsvn.com> wrote:

> On 16 May 2015 at 07:48, Stefan Fuhrmann <stefan.fuhrmann_at_wandisco.com>
> wrote:
> > On Fri, May 15, 2015 at 7:25 PM, Philip Martin <
> philip.martin_at_wandisco.com>
> > wrote:
> >> Another issue: find_entry() now calls drop_entry() in more cases and can
> >> now call it when find_empty==FALSE during read operations. On Unix when
> >> using the read-write lock this means the cache gets modified while only
> >> holding a read lock, not a write lock, and that can corrupt the cache.
> >>
> As far I understand find_entry(find_empty == TRUE) requires write-lock
> for cache, while only read-lock is required for find_empty == FALSE.
> Is correct? It would be nice to document this in docstring.
>

Done in r1679866.

-- Stefan^2.
Received on 2015-05-17 16:46:34 CEST

This is an archived mail posted to the Subversion Dev mailing list.