On Thu, 2009-10-29 at 04:00 +0000, Andrew Bennetts wrote:
> I have a fix for this, it's pretty shallow. I'll submit it for review
> for lp:bzr/2.0 shortly.
>
> More interestingly, we *should* have test coverage for this (e.g.
> blackbox.test_log) but because of the global chk_map page cache the
> necessary code path is never hit.
Is that a between-test isolation issue, or a 'self.run_bzr doesn't start
clean' isolation issue? Would making run_bzr clear that cache make the
issue become visible?
On Thu, 2009-10-29 at 04:00 +0000, Andrew Bennetts wrote:
> I have a fix for this, it's pretty shallow. I'll submit it for review
> for lp:bzr/2.0 shortly.
>
> More interestingly, we *should* have test coverage for this (e.g.
> blackbox.test_log) but because of the global chk_map page cache the
> necessary code path is never hit.
Is that a between-test isolation issue, or a 'self.run_bzr doesn't start
clean' isolation issue? Would making run_bzr clear that cache make the
issue become visible?
-Rob