New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 672119 link

Starred by 1 user

Issue metadata

Status: Verified
Owner:
Closed: Dec 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 1
Type: Bug



Sign in to add a comment

chromeos-base/authpolicy fails to build causing BuildPackages to fail

Project Member Reported by afakhry@chromium.org, Dec 7 2016

Issue description

Multiple PFQ, and informational builders have failed the BuildPackages step because of a failure in building chromeos-base/authpolicy:


authpolicy-0.0.1-r72: [31/36] CXX obj/authpolicy/policy/libauthpolicy.preg_parser.o
authpolicy-0.0.1-r72: FAILED: x86_64-cros-linux-gnu-clang++ -MMD -MF obj/authpolicy/policy/libauthpolicy.preg_parser.o.d -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -Iobj/authpolicy/libauthpolicy.gen/include -Igen/include -I/mnt/host/source/src/platform2 -I/mnt/host/source/src/platform -I/build/peppy/usr/include -Igen -Wall -Wno-psabi -ggdb3 -fstack-protector-strong -Wformat=2 -fvisibility=internal -Wa,--noexecstack -Werror --sysroot=/build/peppy -DDBUS_API_SUBJECT_TO_CHANGE -DUSE_RTTI_FOR_TYPE_TAGS -Wno-c++11-extensions -Wno-unused-local-typedefs -DBASE_VER=395517 -pthread -I/build/peppy/usr/include/dbus-c++-1 -I/build/peppy/usr/include/chromeos -I/build/peppy/usr/include/base-395517 -I/build/peppy/usr/include/glib-2.0 -I/build/peppy/usr/lib64/glib-2.0/include -I/build/peppy/usr/include/dbus-1.0 -I/build/peppy/usr/lib64/dbus-1.0/include -I/build/peppy/usr/include/nss -I/build/peppy/usr/include/nspr -fPIE -std=gnu++11 -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -O2 -pipe -O2 -pipe -O2 -pipe -march=corei7 -g -fno-exceptions -fno-unwind-tables -fno-asynchronous-unwind-tables -clang-syntax -clang-syntax -fno-exceptions -fno-unwind-tables -fno-asynchronous-unwind-tables  -c ../../../../../../../../../mnt/host/source/src/platform2/authpolicy/policy/preg_parser.cc -o obj/authpolicy/policy/libauthpolicy.preg_parser.o
authpolicy-0.0.1-r72: ../../../../../../../../../mnt/host/source/src/platform2/authpolicy/policy/preg_parser.cc:106:15: error: comparison of constant 9223372036854775807 with expression of type 'uint32_t' (aka 'unsigned int') is always true [-Werror,-Wtautological-constant-out-of-range-compare]
authpolicy-0.0.1-r72:   DCHECK(size <= std::numeric_limits<ptrdiff_t>::max());
authpolicy-0.0.1-r72:          ~~~~ ^  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
authpolicy-0.0.1-r72: ../../../../../../../usr/include/base-395517/base/logging.h:727:54: note: expanded from macro 'DCHECK'
authpolicy-0.0.1-r72:   LAZY_STREAM(LOG_STREAM(DCHECK), DCHECK_IS_ON() ? !(condition) : false) \
authpolicy-0.0.1-r72:                                                      ^~~~~~~~~
authpolicy-0.0.1-r72: ../../../../../../../usr/include/base-395517/base/logging.h:389:5: note: expanded from macro 'LAZY_STREAM'
authpolicy-0.0.1-r72:   !(condition) ? (void) 0 : ::logging::LogMessageVoidify() & (stream)
authpolicy-0.0.1-r72:     ^~~~~~~~~
authpolicy-0.0.1-r72: 1 error generated.


Example failed PFQ: https://uberchromegw.corp.google.com/i/chromeos/builders/peppy-chrome-pfq/builds/2841
Example failed informational builder: https://uberchromegw.corp.google.com/i/chromeos.chrome/builders/tricky-tot-chrome-pfq-informational/builds/2799


This CL seems to be the culprit: https://chromium-review.googlesource.com/#/c/416068/. Lutz, can you take a look?
 
Status: Started (was: Assigned)
Fix is already in review, see https://chromium-review.googlesource.com/#/c/417841/. I'll change the CL to point to this bug.
Project Member

Comment 2 by bugdroid1@chromium.org, Dec 7 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromiumos/platform2/+/cd6d46a80cb399ad55f34d9599078b9182e4108e

commit cd6d46a80cb399ad55f34d9599078b9182e4108e
Author: Lutz Justen <ljusten@chromium.org>
Date: Wed Dec 07 16:35:10 2016

authpolicy: Fix 64 bit compilation issues.

Removed a faulty DCHECK.

BUG= chromium:672119 
TEST=Compiles on x86-generic AND on amd64-generic

Change-Id: I39a9776b090e1fcd7bd2304840f72b45dd73f485
Reviewed-on: https://chromium-review.googlesource.com/417841
Reviewed-by: Achuith Bhandarkar <achuith@chromium.org>
Reviewed-by: Ahmed Fakhry <afakhry@chromium.org>
Tested-by: Ahmed Fakhry <afakhry@chromium.org>
Tested-by: Lutz Justen <ljusten@chromium.org>
Commit-Queue: Lutz Justen <ljusten@chromium.org>
Trybot-Ready: Lutz Justen <ljusten@chromium.org>

[modify] https://crrev.com/cd6d46a80cb399ad55f34d9599078b9182e4108e/authpolicy/policy/preg_parser.cc

I've reverted this CL in order to allow the PFQ to move forward:
    https://chromium-review.googlesource.com/#/c/416049/

Assuming that the earlier changes actually fix the compilation
errors, then the reverted CL can be resubmitted once there's a
green CQ run.  Probably.

I would hate to break the build again, so I'm going to add authpolicy to the USE flags in make.defaults for x86 and amd64. Once the build works and is stable, I'll move it back to the chromeos-chrome ebuild.

In a nutshell, the problem with changing the chromeos-chrome-9999.ebuild is:
- My change doesn't uprev chromeos-chrome ebuild, so when my change lands, it doesn't actually get compiled!
- Later, something unrelated uprevs chromeos-chrome and the build breaks.
See https://bugs.chromium.org/p/chromium/issues/detail?id=670430 for more details. Changing make.defaults will force it to be picked up right away and I hpe to get feedback in the precommit queue already.
Status: Fixed (was: Started)
This issue is fixed. See https://chromium-review.googlesource.com/#/c/418200/ for a follow-up.

Comment 6 by dchan@google.com, Mar 4 2017

Labels: VerifyIn-58

Comment 7 by dchan@google.com, Apr 17 2017

Labels: VerifyIn-59

Comment 8 by dchan@google.com, May 30 2017

Labels: VerifyIn-60

Comment 9 by dchan@chromium.org, Aug 1 2017

Labels: VerifyIn-61
Status: Verified (was: Fixed)
Closing. Please reopen it if its not fixed. Thanks!

Sign in to add a comment