Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On Tue, Oct 8, 2013 at 4:35 PM, Andrew Deason wrote: > It really seems like this is going to keep coming up again and again > unless we have a way to generate the .c and .h files separately. That > either means modifying our compile_et to only generate one at once, or > fiddling around with tempor

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Tue, 8 Oct 2013 14:20:38 -0500 Andrew Deason wrote: > cmd_errors.c cmd.h: cmd_errors.et cmd.p.h > $(RM) -f cmd.h cmd_errors.c > echo sleeping ; sleep 1 ; echo done sleeping > ${COMPILE_ET} -p ${srcdir} cmd_errors -h cmd > touch cmd.h > ls > cmd.h: cmd_er

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Tue, 8 Oct 2013 15:38:06 -0400 Michael Meffie wrote: > > Before I go any further I'd also just like to mention that just > > modifying our compile_et to just generate one file at a time would > > seem to be a pretty easy way to get rid of this whole problem. While > > I would like to understan

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On Tue, 8 Oct 2013 14:20:38 -0500 Andrew Deason wrote: > On Tue, 10 Sep 2013 11:02:54 -0400 > Michael Meffie wrote: > > > Russ, thank you for the excellent information. Based on that automake > > manual section, I've pushed a commit to gerrit to fix all the makefile > > rules for compile_et in

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Tue, 10 Sep 2013 11:02:54 -0400 Michael Meffie wrote: > Russ, thank you for the excellent information. Based on that automake > manual section, I've pushed a commit to gerrit to fix all the makefile > rules for compile_et in the tree. > > http://gerrit.openafs.org/#change,10237 This has com

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

Russ Allbery wrote: > This is the "command that creates two files" make problem, which is a > well-known problem with a wide variety of solutions with, if not perfect > behavior, at least well-understood properties. > > info automake faq 'multiple outputs' > > will explain a bunch of the po

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 05 Sep 2013 21:49:31 -0400 Jason Edgecombe wrote: > After some experimentation, I've changed the old git clean statement > on the debian slave from: > git clean -X -f -e !.buildbot-sourcedata > to: > git clean -d -x -f -e '.buildbot-sourcedata' > git clean -d -X -f -e '!.bu

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

Michael Meffie writes: > It seems to me, a more proper fix for all this would be to have a target > for bnode.h and a target for boserr.c which are independent (and the > same for all the other placed compile_et is run). I was wondering how > krb5 copes with compile_et, and that is basically wha

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 5 Sep 2013 11:17:27 -0400 Benjamin Kaduk wrote: > On Thu, 5 Sep 2013, Andrew Deason wrote: > > > On Thu, 5 Sep 2013 11:07:23 -0400 > > Michael Meffie wrote: > > > >> I've seen the same thing happen on debian by just doing a 'make clean' > >> in, for example, src/bozo. I recall changing

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On 09/06/2013 03:28 AM, Christof Hanke wrote: On Thu, 05 Sep 2013 21:56:21 -0400 Jason Edgecombe wrote: On 09/05/2013 12:19 PM, Jeffrey Altman wrote: On 9/5/2013 12:08 PM, Andrew Deason wrote: On Thu, 5 Sep 2013 09:58:23 -0500 Andrew Deason wrote: I'm not sure if it's why this is happenin

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 05 Sep 2013 21:56:21 -0400 Jason Edgecombe wrote: > On 09/05/2013 12:19 PM, Jeffrey Altman wrote: > > On 9/5/2013 12:08 PM, Andrew Deason wrote: > >> On Thu, 5 Sep 2013 09:58:23 -0500 > >> Andrew Deason wrote: > >> > >>> I'm not sure if it's why this is happening, but the builders don't

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On 09/05/2013 12:19 PM, Jeffrey Altman wrote: On 9/5/2013 12:08 PM, Andrew Deason wrote: On Thu, 5 Sep 2013 09:58:23 -0500 Andrew Deason wrote: I'm not sure if it's why this is happening, but the builders don't seem to be removing all 'cruft' between build runs, since they are running 'git cl

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On 09/05/2013 12:19 PM, Jeffrey Altman wrote: On 9/5/2013 12:08 PM, Andrew Deason wrote: On Thu, 5 Sep 2013 09:58:23 -0500 Andrew Deason wrote: I'm not sure if it's why this is happening, but the builders don't seem to be removing all 'cruft' between build runs, since they are running 'git cl

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 05 Sep 2013 12:18:19 -0400 Jeffrey Altman wrote: > My point is that the rule that was selected was set on a per builder > basis because of the different revisions on each build slave. If the > revisions of buildbot and/or git changed without the build rules being > updated, the breakage

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

all the patches in gerrit to fix it so far have been integrated :) On Thu, Sep 5, 2013 at 11:22 AM, Harald Barth wrote: > > It's actually a shame that the dependencies and rules in the makefile > are not good enough so that we are forced to first clean the tree and > then build 99% again (mostl

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 5 Sep 2013 09:58:23 -0500 Andrew Deason wrote: > I'm not sure if it's why this is happening, but the builders don't seem > to be removing all 'cruft' between build runs, since they are running > 'git clean -X' instead of 'git clean -x'. In particular on that slave, > the 'include' directo

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 5 Sep 2013 09:58:23 -0500 Andrew Deason wrote: > On Thu, 5 Sep 2013 10:04:33 +0200 > Christof Hanke wrote: > > > Including ./rxkad.p.h at beginning of rxkad.h file. > > /usr/bin/install -c -m 644 > > /home/buildslave/buildbot/slave/openafs/full/build/include/rx/rxkad.h > > /usr/bin/in

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 05 Sep 2013 11:34:33 -0400 Jeffrey Altman wrote: > I doubt that anyone has been tracking the git and buildbot software > revisions as builders have been upgraded and replaced. We know what version of buildbot the slaves are running from here:

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 05 Sep 2013 17:22:41 +0200 (CEST) Harald Barth wrote: > It's actually a shame that the dependencies and rules in the makefile > are not good enough so that we are forced to first clean the tree and > then build 99% again (mostly exactly as the previous run) besides the > few files which a

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 5 Sep 2013 11:07:23 -0400 Michael Meffie wrote: > I've seen the same thing happen on debian by just doing a 'make clean' > in, for example, src/bozo. I recall changing the makefile as Ben says > to replace the $? with the actual filename, but wasn't sure if that > was the right way to fi

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On 9/5/13 11:07, "Michael Meffie" wrote: >On Thu, 5 Sep 2013 09:58:23 -0500 >Andrew Deason wrote: > >> On Thu, 5 Sep 2013 10:04:33 +0200 >> Christof Hanke wrote: >> >> > Including ./rxkad.p.h at beginning of rxkad.h file. >> > /usr/bin/install -c -m 644 >>/home/buildslave/buildbot/slave/openaf

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 5 Sep 2013 10:04:33 +0200 Christof Hanke wrote: > Including ./rxkad.p.h at beginning of rxkad.h file. > /usr/bin/install -c -m 644 > /home/buildslave/buildbot/slave/openafs/full/build/include/rx/rxkad.h > /usr/bin/install: missing destination file operand after > `/home/buildslave/buil

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On 9/5/2013 12:08 PM, Andrew Deason wrote: > On Thu, 5 Sep 2013 09:58:23 -0500 > Andrew Deason wrote: > >> I'm not sure if it's why this is happening, but the builders don't seem >> to be removing all 'cruft' between build runs, since they are running >> 'git clean -X' instead of 'git clean -x'.

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On 9/5/2013 11:53 AM, Andrew Deason wrote: > On Thu, 05 Sep 2013 11:34:33 -0400 > Jeffrey Altman wrote: > >> I doubt that anyone has been tracking the git and buildbot software >> revisions as builders have been upgraded and replaced. > > We know what version of buildbot the slaves are running f

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On 9/5/2013 10:58 AM, Andrew Deason wrote: > I'm not sure if it's why this is happening, but the builders don't seem > to be removing all 'cruft' between build runs, since they are running > 'git clean -X' instead of 'git clean -x'. In particular on that slave, > the 'include' directory is staying

[OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 5 Sep 2013 11:17:27 -0400 (EDT) Benjamin Kaduk wrote: > I think I remember something about buildbot configuration files or > cache or something that we did not want to remove. That discussion was about .buildbot.sourcedata, which we are currently avoiding deleting by specifying '-e !.bui

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

It's actually a shame that the dependencies and rules in the makefile are not good enough so that we are forced to first clean the tree and then build 99% again (mostly exactly as the previous run) besides the few files which actually were touched by the latest patch. Waste of CPU cycles, isn't it

Re: [OpenAFS-devel] Re: debian-linux-i386-builder issues

On Thu, 5 Sep 2013, Andrew Deason wrote: On Thu, 5 Sep 2013 11:07:23 -0400 Michael Meffie wrote: I've seen the same thing happen on debian by just doing a 'make clean' in, for example, src/bozo. I recall changing the makefile as Ben says to replace the $? with the actual filename, but wasn't