coreboot should not try to clone git repos |
||||
Issue descriptionI happened to notice a trybot build flake: https://cros-goldeneye.corp.google.com/chromeos/healthmonitoring/buildDetails?buildbucketId=8928587417136056976 coreboot-0.0.1-r2812: Cloning SeaBIOS from Git coreboot-0.0.1-r2812: Cloning into 'seabios'... ... coreboot-0.0.1-r2812: fatal: unable to access 'https://review.coreboot.org/seabios.git/': Couldn't resolve host 'review.coreboot.org' ... coreboot-0.0.1-r2812: make[1]: *** [Makefile:19: seabios] Error 128 coreboot-0.0.1-r2812: make: *** [payloads/external/Makefile.inc:74: payloads/external/SeaBIOS/seabios/out/bios.bin.elf] Error 2 coreboot-0.0.1-r2812: make: *** Waiting for unfinished jobs.... payloads/external/SeaBIOS/Makefile contains this: project_git_repo=https://review.coreboot.org/seabios.git ... seabios: echo " Cloning SeaBIOS from Git" git clone $(project_git_repo) $(project_dir) That's a big no-no. I'd think portage should be preventing network access like that too...
,
Nov 28
Hmm, is it possible this is only a problem with Slippy? I was indiscriminately launching all available pre-cq builders, to get better build coverage. But it looks like Slippy isn't really used anywhere -- and it only has a private overlay. Perhaps Slippy should be killed entirely?
,
Nov 28
we can kill slippy. it was a ref board and i don't think we have any active configs for it anymore. Bernie can confirm.
,
Nov 28
I think this would do it. https://chrome-internal-review.googlesource.com/c/chromeos/manifest-internal/+/722323 https://chromium-review.googlesource.com/c/chromiumos/chromite/+/1354385 Might be worth getting some analysis from FW team still, to make sure this is just a corner case in an unmaintained overlay, and not something we expect to crop up again accidentally.
,
Nov 28
You can kill slippy.
,
Nov 29
coreboot's default config is to build seabios (and for that, to download its git repo). None of our properly configured builds should ever run into that though. In https://chromium-review.googlesource.com/600771 we removed slippy's config, so I guess it picked the defaults instead. That it only occurred now (the commit is from early August) may be a testament to how long this buildbot config hasn't been used.
,
Nov 29
Cool, thanks for the info. So I'll just shop for reviews on the "how" (above CLs) of killing Slippy.
,
Nov 29
Removing slippy SGTM, we don't use it on ToT for normal builds.
,
Dec 4
The following revision refers to this bug: https://chromium.googlesource.com/chromiumos/chromite/+/24e1da6ab5ed606a4ecf4a1de96ff21b3312d74a commit 24e1da6ab5ed606a4ecf4a1de96ff21b3312d74a Author: Brian Norris <briannorris@chromium.org> Date: Tue Dec 04 08:11:36 2018 config: kill slippy It's just a reference board, has no public overlay, is not used anywhere, and doesn't even build correctly any more. BUG= chromium:909874 TEST=precq Change-Id: Ic3b401e99eafa8bd16bb49a6880b116df1bc2f6e Signed-off-by: Brian Norris <briannorris@chromium.org> Reviewed-on: https://chromium-review.googlesource.com/1354385 Commit-Ready: ChromeOS CL Exonerator Bot <chromiumos-cl-exonerator@appspot.gserviceaccount.com> [modify] https://crrev.com/24e1da6ab5ed606a4ecf4a1de96ff21b3312d74a/config/chromeos_config_boards.py [modify] https://crrev.com/24e1da6ab5ed606a4ecf4a1de96ff21b3312d74a/lib/paygen/testdata/paygen.json [modify] https://crrev.com/24e1da6ab5ed606a4ecf4a1de96ff21b3312d74a/config/config_dump.json [modify] https://crrev.com/24e1da6ab5ed606a4ecf4a1de96ff21b3312d74a/lib/paygen/paygen_build_lib_unittest.py [modify] https://crrev.com/24e1da6ab5ed606a4ecf4a1de96ff21b3312d74a/config/waterfall_layout_dump.txt
,
Dec 5
The following revision refers to this bug: https://chrome-internal.googlesource.com/chromeos/manifest-internal/+/4f49ec495f0674440ea1203667c81ba222a79e4f commit 4f49ec495f0674440ea1203667c81ba222a79e4f Author: Brian Norris <briannorris@chromium.org> Date: Wed Dec 05 01:40:16 2018
,
Dec 5
|
||||
►
Sign in to add a comment |
||||
Comment 1 by vapier@chromium.org
, Nov 28Summary: coreboot should not try to clone git repos (was: Builds should not depend on resolveing review.coreboot.org)