Re: [PATCH v7 2/2] Verify index file before we opportunistically update it

From: Yiannis Marangos <yiannis.marangos@gmail.com>
Date: Apr 11, 2014, 15:58 UTC
Message-ID: <20140411155857.GA12742@abyss.hitronhub.home>
In-Reply-To: <53479DFD.4020702@web.de>

On Fri, Apr 11, 2014 at 09:47:09AM +0200, Torsten Bögershausen wrote:

Show 5 quoted lines

>   6. process A applies a commit:
>       - read the index into memory
>       - take the lock
>       - update the index file on disc
>       - release the lock

So here we can have race condition. Maybe we should implement Duy's idea of verifying the sha1 in write_index()?

Show 8 quoted lines

> The new code works like this:
>
>   8. process B applies a commit:
>       - take the lock
>       - verifies tha the index file on disc has the same sha as the one read before
>       # And if not: What do we do? die() or retry() ?
>       - update the index file on disc
>       - release the lock

IMO, we must never die() if we want to do opportunistic update and we should never retry because it's a bit expensive if we do opportunistic update.

-- Yiannis