Hi Ken,
----- Original Message -----
> [...]
> I think I've dealt with most of the VM environmental failures now, so these
> are probably real failures.
>
> 382 looks like a recent regression/change
Unhelpfully, passing reliably here.
> 666 remains a royal pain in the backside
I've attached a patch that implements a suggestion from Mark Wielaard to
help this (well-numbered test from hell) with older versions of valgrind.
>From the posted .bad files, this might help "fix" some failure cases.
> 544 ditto
This'll likely be from recent changes in this area from Mark (collectl2pcp)
though its passing everywhere here, probably best to wait for his insight.
> 651 ditto
The only clue I have here so far is that 651 only fails for me with a non-
secure-sockets enabled build. Still investigating though.
> 365 fixed, but not yet propagated thru all of the QA farm
> ...
> Builds still broken on some platforms.
Hopefully those two issues are resolved everywhere now.
> grundy.sgi.com is out of the pool as pmmgr regularly kills the machine (this
> means hang, pmmgr spinning, cannot kill, cannot reboot, power cycle is the
> only option to recover).
(memory leak? OOM might at least partially explain these observations)
cheers.
--
Nathan
valgrind.patch
Description: Text Data
|