|
|
Chromium Code Reviews|
Created:
7 years, 6 months ago by robliao Modified:
7 years, 6 months ago CC:
chromium-reviews, vadimt Base URL:
https://src.chromium.org/chrome/trunk/src/ Visibility:
Public. |
DescriptionFix Landmines MSVS Version Checking
GYP_MSVS_VERSION is not always a number (e.g. it can be 2010e),
causing the int conversion to fail.
BUG=249378
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=206253
Patch Set 1 #
Total comments: 2
Patch Set 2 : Quick CR Feedback #
Total comments: 2
Messages
Total messages: 14 (0 generated)
Hi Robbie, My Windows build just broke due to a GYP_MSVS_VERSION compare. GYP_MSVS_VERSION can be 2010e (like on my machine), causing a break. Let me know if you'd like to submit delta (alternatively, I can do it if you want). Thanks! Robert
On 2013/06/13 17:38:29, Robert Liao wrote: > Hi Robbie, > > My Windows build just broke due to a GYP_MSVS_VERSION compare. GYP_MSVS_VERSION > can be 2010e (like on my machine), causing a break. > > Let me know if you'd like to submit delta (alternatively, I can do it if you > want). > > Thanks! > > Robert lgtm! Sorry for that... I didn't realize that this could be a string!
ben: Please provide owner approval for this build fix.
https://codereview.chromium.org/16978002/diff/1/build/landmines.py File build/landmines.py (right): https://codereview.chromium.org/16978002/diff/1/build/landmines.py#newcode72 build/landmines.py:72: return val if val else None Actually, you can just make this return os.environ.get('GYP_MSVS_VERSION', None)
Fix Landmines MSVS Version Checking GYP_MSVS_VERSION is not always a number (e.g. it can be 2010e), causing the int conversion to fail.
https://codereview.chromium.org/16978002/diff/1/build/landmines.py File build/landmines.py (right): https://codereview.chromium.org/16978002/diff/1/build/landmines.py#newcode72 build/landmines.py:72: return val if val else None On 2013/06/13 18:12:44, iannucci wrote: > Actually, you can just make this > > return os.environ.get('GYP_MSVS_VERSION', None) Done.
https://codereview.chromium.org/16978002/diff/7002/build/landmines.py File build/landmines.py (right): https://codereview.chromium.org/16978002/diff/7002/build/landmines.py#newcode156 build/landmines.py:156: gyp_msvs_version() == '2012' and this should be gyp_msvs_version() in ('2012', '2012e') i guess?
On 2013/06/13 18:28:57, scottmg wrote: > https://codereview.chromium.org/16978002/diff/7002/build/landmines.py > File build/landmines.py (right): > > https://codereview.chromium.org/16978002/diff/7002/build/landmines.py#newcode156 > build/landmines.py:156: gyp_msvs_version() == '2012' and > this should be gyp_msvs_version() in ('2012', '2012e') i guess? Could be, but we really only care about nuking the 2012 trybots. No reason to nuke the 2012e ones.
https://codereview.chromium.org/16978002/diff/7002/build/landmines.py File build/landmines.py (right): https://codereview.chromium.org/16978002/diff/7002/build/landmines.py#newcode156 build/landmines.py:156: gyp_msvs_version() == '2012' and Propagating iannucci's comment: Could be, but we really only care about nuking the 2012 trybots. No reason to nuke the 2012e ones. On 2013/06/13 18:28:57, scottmg wrote: > this should be gyp_msvs_version() in ('2012', '2012e') i guess?
there's really no OWNER in src/build? I'm not a good person to review this kind of change.
On 2013/06/13 19:43:06, Ben Goodger (Google) wrote: > there's really no OWNER in src/build? I'm not a good person to review this kind > of change. TBH, I originally wrote landmines.py, so I know the code decently well :)
build/OWNERS is *, iannucci should have been enough, but lgtm anyway.
CQ is trying da patch. Follow status at https://chromium-status.appspot.com/cq/robliao@chromium.org/16978002/7002
Message was sent while issue was closed.
Change committed as 206253 |
