Re: VM: Fix nasty and subtle race in shared mmap'ed page writeback

[Andrea Gelmini <gelma@gelma.net>]

On Fri, Jan 05, 2007 at 04:36:06PM +1100, Nick Piggin wrote:
> I was thinking like a 100 line C program that I can reproduce here ;)
not soon, but I'll try to produce it.

> If you can even describe the steps it does: (eg. mmap file A, write(2) to
> it, truncate it, ...., should contain 1s but it contains 0s!), then we
> might have some suggestions to try.
I ignore this... I mean, BerkelyDB does all this on its own (we just
talk with the library). we never touch directly the DB.

> One obvious thing is change filesystems or filesystem block sizes, try
> ext2 or even tmpfs. Another is to try using write(2) instead of mmap to
> write data.
well, I tried all filesystems, tmpfs also. I know BDB use a lot mmap.

thanks,
gelma