ninja , compile chromium error ! help ! midl.exe output different from files in gen/chrome/elevation_service, see c:\users\worre\appdata\local\temp\tmpga1ovn

485 views
Skip to first unread message

Y Z

unread,
Oct 22, 2019, 1:32:09 AM10/22/19
to Chromium-dev
My Errors:
E:\25.chrome_source\chromium\src>ninja -C ../out/release_x64 chrome
ninja: Entering directory `../out/release_x64'
[34/21828] ACTION //chrome/elevation_service:elevation_service_idl_idl_action(//build/toolchain/win:win_clang_x64)
FAILED: gen/chrome/elevation_service/elevation_service_idl.h gen/chrome/elevation_service/elevation_service_idl.dlldata.c gen/chrome/elevation_service/elevation_service_idl_i.c gen/chrome/elevation_service/elevation_service_idl_p.c gen/chrome/elevation_service/elevation_service_idl.tlb
E:/25.chrome_source/depot_tools/bootstrap-3_8_0b1_chromium_1_bin/python/bin/python.exe ../../src/build/toolchain/win/midl.py environment.x64 E:/25.chrome_source/chromium/src/third_party/win_build_output/midl/chrome/elevation_service gen/chrome/elevation_service none elevation_service_idl.tlb elevation_service_idl.h elevation_service_idl.dlldata.c elevation_service_idl_i.c elevation_service_idl_p.c ../../src/chrome/elevation_service/elevation_service_idl.idl /char signed /env x64 /Oicf
midl.exe output different from files in gen/chrome/elevation_service, see c:\users\worre\appdata\local\temp\tmpga1ovn
--- gen/chrome/elevation_service\elevation_service_idl_p.c
+++ c:\users\worre\appdata\local\temp\tmpga1ovn\elevation_service_idl_p.c
@@ -6,7 +6,7 @@
  /* File created by MIDL compiler version 8.xx.xxxx */
 /* at a redacted point in time
  */
-/* Compiler settings for ../../chrome/elevation_service/elevation_service_idl.idl:
+/* Compiler settings for ../../src/chrome/elevation_service/elevation_service_idl.idl:
     Oicf, W1, Zp8, env=Win64 (32b run), target_arch=AMD64 8.xx.xxxx
     protocol : dce , ms_ext, c_ext, robust
     error checks: allocation ref bounds_check enum stub_data

--- gen/chrome/elevation_service\elevation_service_idl.h
+++ c:\users\worre\appdata\local\temp\tmpga1ovn\elevation_service_idl.h
@@ -6,7 +6,7 @@
  /* File created by MIDL compiler version 8.xx.xxxx */
 /* at a redacted point in time
  */
-/* Compiler settings for ../../chrome/elevation_service/elevation_service_idl.idl:
+/* Compiler settings for ../../src/chrome/elevation_service/elevation_service_idl.idl:
     Oicf, W1, Zp8, env=Win64 (32b run), target_arch=AMD64 8.xx.xxxx
     protocol : dce , ms_ext, c_ext, robust
     error checks: allocation ref bounds_check enum stub_data

--- gen/chrome/elevation_service\elevation_service_idl_i.c
+++ c:\users\worre\appdata\local\temp\tmpga1ovn\elevation_service_idl_i.c
@@ -8,7 +8,7 @@
  /* File created by MIDL compiler version 8.xx.xxxx */
 /* at a redacted point in time
  */
-/* Compiler settings for ../../chrome/elevation_service/elevation_service_idl.idl:
+/* Compiler settings for ../../src/chrome/elevation_service/elevation_service_idl.idl:
     Oicf, W1, Zp8, env=Win64 (32b run), target_arch=AMD64 8.xx.xxxx
     protocol : dce , ms_ext, c_ext, robust
     error checks: allocation ref bounds_check enum stub_data

To rebaseline:
  copy /y c:\users\worre\appdata\local\temp\tmpga1ovn\* E:\25.chrome_source\chromium\src\third_party\win_build_output\midl\chrome\elevation_service\x64
[43/21828] CXX obj/components/blacklist/opt_out_blacklist/opt_out_blacklist/opt_out_blacklist.obj
ninja: build stopped: subcommand failed.

I have try vs2017 and vs2019, but still failed, why?



Pavel Seleznev

unread,
Oct 23, 2019, 11:43:58 AM10/23/19
to Chromium-dev
Hi,

I have the same error during building all projects (78.0.3904.70), but I didn't got it if I build only chromium

Regards,
Pavel

Bruce Dawson

unread,
Oct 23, 2019, 1:46:13 PM10/23/19
to Chromium-dev
It looks like the midl compilation checks have an embedded assumption that out is inside of src whereas your out directory is a child directory to src. Is this a change that you have made recently?

I just synced to latest and confirmed that this command works:

ninja -C out\release_64 gen/chrome/elevation_service/elevation_service_idl.tlb

but this command fails:

ninja -C ..\out\release_64 gen/chrome/elevation_service/elevation_service_idl.tlb

So, your simplest solution is to move your out directory to be inside src instead of beside it. If you think that having out beside src is important then consider filing a bug, and updating the comparison script to ignore these differences.

Bruce Dawson

unread,
Oct 23, 2019, 2:03:49 PM10/23/19
to Chromium-dev
I said "your out directory is a child directory to src" but I meant that your out directory is a sibling to src, instead of being a child.

chromium\src\out\release_64 - good

chromium\src
                \out\release_64 - bad (or, at least, causing these problems)

--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev
---
You received this message because you are subscribed to a topic in the Google Groups "Chromium-dev" group.
To unsubscribe from this topic, visit https://groups.google.com/a/chromium.org/d/topic/chromium-dev/Iac7UjZJnj4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to chromium-dev...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/ae696e5d-3d18-497d-88d8-234969d2ab26%40chromium.org.

Dirk Pranke

unread,
Oct 23, 2019, 2:11:59 PM10/23/19
to Bruce Dawson, Chromium-dev
As a side note, I'd actually be quite surprised if many things didn't break if your output directory was not inside src. In theory it could work, but we have no continuous builds that test this and I don't even know of any individuals who try it.

It would be interesting to see how hard it would be to make it work.

-- Dirk

You received this message because you are subscribed to the Google Groups "Chromium-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chromium-dev...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/CAE5mQiOWBN%2BpUHbZB9LQ%2Bxjn_TTiXsGAoaVgN24RZc-M_SLOFw%40mail.gmail.com.

Mostyn Bramley-Moore

unread,
Oct 23, 2019, 3:52:05 PM10/23/19
to Dirk Pranke, Bruce Dawson, Chromium-dev
On Wed, Oct 23, 2019 at 8:11 PM Dirk Pranke <dpr...@chromium.org> wrote:
As a side note, I'd actually be quite surprised if many things didn't break if your output directory was not inside src. In theory it could work, but we have no continuous builds that test this and I don't even know of any individuals who try it.

It would be interesting to see how hard it would be to make it work.

-- Dirk

IIRC the bot swarming setup assumes that the output directory is exactly two levels deep inside the source root.  So even if you get local builds with the output directory in an unusual place, the official bots are likely to have trouble.

-Mostyn.
 

Dirk Pranke

unread,
Oct 23, 2019, 3:58:14 PM10/23/19
to Mostyn Bramley-Moore, Bruce Dawson, Chromium-dev
On Wed, Oct 23, 2019 at 12:51 PM Mostyn Bramley-Moore <mos...@vewd.com> wrote:
On Wed, Oct 23, 2019 at 8:11 PM Dirk Pranke <dpr...@chromium.org> wrote:
As a side note, I'd actually be quite surprised if many things didn't break if your output directory was not inside src. In theory it could work, but we have no continuous builds that test this and I don't even know of any individuals who try it.

It would be interesting to see how hard it would be to make it work.

-- Dirk

IIRC the bot swarming setup assumes that the output directory is exactly two levels deep inside the source root.  So even if you get local builds with the output directory in an unusual place, the official bots are likely to have trouble.

All of the bots use out*/*/, yes, so anything that broke that would likely not land or would be quickly reverted. But, that's different from saying that something else *shouldn't* work. We don't intentionally explicitly require something two levels deep, it's just that there are likely to be bugs with anything else.

-- Dirk

Nico Weber

unread,
Oct 23, 2019, 6:26:12 PM10/23/19
to Mostyn Bramley-Moore, Dirk Pranke, Bruce Dawson, Chromium-dev
In addition to that, goma's cache sharing also only works if the path from build dir to src dir is ../..

Pavel Seleznev

unread,
Oct 24, 2019, 5:45:08 AM10/24/19
to Chromium-dev
Hello,

Not sure that I understand you correctly, but I am inside the src directory, here is my error log output (I was formatted it a bit)

G:\autobuild\clean78.0.3904.70\src>ninja -C out/Default all ninja: Entering directory `out/Default'

[725/35231] ACTION
//chrome/browser/browser_switcher/bho:ie_bho_idl_idl_action(//build/toolchain/win:win_clang_x86)
FAILED:
gen/chrome/browser/browser_switcher/bho/ie_bho_idl.h
gen/chrome/browser/browser_switcher/bho/ie_bho_idl.dlldata.c
gen/chrome/browser/browser_switcher/bho/ie_bho_idl_i.c
gen/chrome/browser/browser_switcher/bho/ie_bho_idl_p.c
G:/depot_tools/bootstrap-3_8_0b1_chromium_1_bin/python/bin/python.exe
../../build/toolchain/win/midl.py environment.x86
G:/autobuild/clean78.0.3904.70/src/third_party/win_build_output/midl/chrome/browser/browser_switcher/bhogen/chrome/browser/browser_switcher/bho
n one ie_bho_idl.tlb ie_bho_idl.h ie_bho_idl.dlldata.c ie_bho_idl_i.c ie_bho_idl_p.c
../../chrome/browser/browser_switcher/bho/ie_bho_i dl.idl /char signed /env
win32 /Oicf midl.exe output different from files in gen/chrome/browser/browser_switcher/bho, see
c:\users\p3dcd~1.sel\appdata\local\temp\tmpuw2ga s To rebaseline: copy /y
c:\users\p3dcd~1.sel\appdata\local\temp\tmpuw2gas\*
G:\autobuild\clean78.0.3904.70\src\third_party\win_build_output\midl\chrome\browser\browser_switcher\bho\x86

Bruce Dawson

unread,
Oct 24, 2019, 12:22:37 PM10/24/19
to pavel.s...@gmail.com, Chromium-dev
The errors you are seeing appear to be unrelated to the original poster's issues. I can't tell from the output what is going on. These steps are working for other people so there is probably something unusual about your environment (VC++ version, gn args, etc.) that is causing the discrepancy. You'll have to share more information or investigate more in order to understand the failure.

--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev
---
You received this message because you are subscribed to a topic in the Google Groups "Chromium-dev" group.
To unsubscribe from this topic, visit https://groups.google.com/a/chromium.org/d/topic/chromium-dev/Iac7UjZJnj4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to chromium-dev...@chromium.org.
Reply all
Reply to author
Forward
0 new messages