[Scons-users] scons-local-2.5.1 build and clean shows random failure
Bill Deegan
bill at baddogconsulting.com
Tue Jul 25 11:01:44 EDT 2017
I need a taskmaster trace file from a failing build.
A trace from a build which doesn't fail doesn't show the issue.
By default all directories depend on file in that directory.
Try adding the following to this line:
desc = desc + \
" Internal Error: no cycle found for node %s (%s) in state %s\n" % \
(node, repr(node), StateString[node.get_state()])
+ [n.get_abspath() for n in self.pending_children] + nclist
Hopefully that should give useful info about the issue.
-Bill
On Tue, Jul 25, 2017 at 7:53 AM, Hua Yanghao <huayanghao at gmail.com> wrote:
> On Tue, Jul 25, 2017 at 4:49 PM, Bill Deegan <bill at baddogconsulting.com>
> wrote:
> >
> >
> > On Tue, Jul 25, 2017 at 7:45 AM, Hua Yanghao <huayanghao at gmail.com>
> wrote:
> >>
> >> On Tue, Jul 25, 2017 at 4:43 PM, Bill Deegan <bill at baddogconsulting.com
> >
> >> wrote:
> >> > I need a taskmaster trace from a parallel build which fails to be able
> >> > to
> >> > see more about what's happening.
> >>
> >> I log I attached before is from a parallel build!
> >
> > From a failing parallel build?
>
> Sorry that was from a parallel clean. I will try to reproduce a
> failure on parallel build
> (don't remember if the build have the same issue). but isn't a
> parallel clean even
> a simpler case and easier to debug?
>
> >>
> >> > A single thread build after a failed build doesn't yield any useful
> >> > information.
> >> >
> >> > As a data point this issue is not that widespread as parallel build is
> >> > used
> >> > widely without issue.
> >>
> >> I will either try to fix this issue in scons or I will abandon it :-)
> >> I hope you could help me as well :-)
> >>
> >> > Are you using SideEffects()?
> >>
> >> Yes I am using it.
> >
> >
> > Are there any SideEffects pointing at directories and not files?
> > Or pointing into the directory indicated in the non loop?
> >
> It is pointing to a output folder:
> output/linux64_full/linux64_full.map
> (which is a generated memory map from linking).
>
> The failure is pointing to the output/linux64_full folder only.
> _______________________________________________
> Scons-users mailing list
> Scons-users at scons.org
> https://pairlist4.pair.net/mailman/listinfo/scons-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist4.pair.net/pipermail/scons-users/attachments/20170725/365e8353/attachment-0001.html>
More information about the Scons-users
mailing list