[dm-devel] Re: possible regression by the barrier patch in 2.6.30-rc2

[Alasdair G Kergon]

On Wed, Nov 04, 2009 at 08:53:23PM -0500, Mikulas Patocka wrote:
> On Wed, 28 Oct 2009, Alasdair G Kergon wrote:
> > This can be either a "suspend with flush" OR the
> > flush could have been issued prior to the suspend (and any decent database
> > would have done that).
 
> I'm somehow starting to think that "flush" suspend is not needed at all 
> and all suspends may be "noflush". 

Indeed - that was what my email was trying to say - 'decent database' is
the one you describe, which waits properly, and the only reason for 'flush
suspend' is backwards compatibility with our previous releases to support
poorly-written software (which quite probably doesn't exist).

Alasdair