Windows build failed (PCH failure?) |
|
Issue descriptionOS: Windows 10 10586 I was building successfully for weeks with the following args.gn: is_win_fastlink = true is_component_build = true target_cpu = "x86" symbol_level = 1 enable_nacl = false This morning, I did a git pull --rebase which brought in this CL: https://chromium.googlesource.com/chromium/src/+/e8de0e46213b60669a73760f30aa59c74f43f1f9%5E%21/#F5 Attempting to rebuild resulted in multiple build failures complaining that the IsSameOriginOrCORSSuccessful() function defined in Resource.h did not exist (despite the function clearly being present in the file). --------- ninja -C out/default ninja: Entering directory `out/default' [1/17641] CXX obj/third_party/WebKit/Source/core/frame/frame/SubresourceIntegrity.obj FAILED: obj/third_party/WebKit/Source/core/frame/frame/SubresourceIntegrity.obj ninja -t msvc -e environment.x86 -- "c:\src\c\src\third_party\depot_tools\win_toolchain\vs_files\f53e4598951162bad6330f7a167486c7ae5db1e5\vc\bin\amd64_x86/cl.exe" /nologo /showIncludes /FC @obj/third_party/WebKit/Source/core/frame/frame/SubresourceIntegrity.obj.rsp /c ../../third_party/WebKit/Source/core/frame/SubresourceIntegrity.cpp /Foobj/third_party/WebKit/Source/core/frame/frame/SubresourceIntegrity.obj /Fd"obj/third_party/WebKit/Source/core/frame/frame_cc.pdb" c:\src\c\src\third_party\webkit\source\core\frame\subresourceintegrity.cpp(133): error C2039: 'Fuzzle': is not a member of 'blink::Resource' c:\src\c\src\third_party\webkit\source\core\frame\subresourceintegrity.h(20): note: see declaration of 'blink::Resource' c:\src\c\src\third_party\webkit\source\core\frame\subresourceintegrity.cpp(134): error C2039: 'IsSameOriginOrCORSSuccessful': is not a member of 'blink::Resource' c:\src\c\src\third_party\webkit\source\core\frame\subresourceintegrity.h(20): note: see declaration of 'blink::Resource' [2/17641] CXX obj/third_party/WebKit/Source/core/css/css/RemoteFontFaceSource.obj FAILED: obj/third_party/WebKit/Source/core/css/css/RemoteFontFaceSource.obj ninja -t msvc -e environment.x86 -- "c:\src\c\src\third_party\depot_tools\win_toolchain\vs_files\f53e4598951162bad6330f7a167486c7ae5db1e5\vc\bin\amd64_x86/cl.exe" /nologo /showIncludes /FC @obj/third_party/WebKit/Source/core/css/css/RemoteFontFaceSource.obj.rsp /c ../../third_party/WebKit/Source/core/css/RemoteFontFaceSource.cpp /Foobj/third_party/WebKit/Source/core/css/css/RemoteFontFaceSource.obj /Fd"obj/third_party/WebKit/Source/core/css/css_cc.pdb" c:\src\c\src\third_party\webkit\source\core\css\remotefontfacesource.cpp(128): error C2039: 'IsSameOriginOrCORSSuccessful': is not a member of 'blink::FontResource' c:\src\c\src\third_party\webkit\source\core\loader\resource\fontresource.h(44): note: see declaration of 'blink::FontResource' c:\src\c\src\third_party\webkit\source\core\css\remotefontfacesource.cpp(340): error C2039: 'IsSameOriginOrCORSSuccessful': is not a member of 'blink::FontResource' c:\src\c\src\third_party\webkit\source\core\loader\resource\fontresource.h(44): note: see declaration of 'blink::FontResource' [50/17641] CXX obj/third_party/WebKit/Source/core/offscreencanvas/offscreencanvas/Precompile-core.cc.obj --------- Modifying resource.h and resource.cc had no effect on these errors. When I deliberately introduced an error in the IsSameOriginOrCORSSuccessful function inside resource.h, it was reported but didn't impact or otherwise clear the existing error claiming that the function did not exist. Unfortunately, my attempts to track down the appropriate PCH/OBJ were not successful and ultimately I ended up running 'ninja -C out/Default -t clean' after which time the build started working again.
,
Jun 28 2017
Also, I highly recommend "git rebase-update". It is like "git pull --rebase" except that it advances all of your branches. This minimizes the differences between branches and therefore makes building after switching between branches much faster.
,
Jun 28 2017
Unfortunately I was not able to reproduce this problem. I synced to 2bf9f75 (one commit before e8de0e4) and built Chrome with your settings (and goma, to save time). Then I synced to e8de0e4 and built again. 4139 build steps were executed and the build of Chrome worked. I tried a second time after fixing my landmines so that I could do a proper gclient sync. gn clean, sync to 2bf9f75, gclient sync, build chrome, sync to e8de0e4, gclient sync, build chrome, 4189 build steps were executed and the build of Chrome worked. So, no repro so far. Also, I don't see any .pch files which seems odd. |
|
►
Sign in to add a comment |
|
Comment 1 by brucedaw...@chromium.org
, Jun 28 2017