RunTests.bat on Windows fails test 8 for me

[eli-schwartz]

I don't know what the cause of this might be. I have a thirdparty port of pcre2 to the Meson build system (for embedding into other projects using meson) and as part of the port we try to run the testsuite.

For reference, here is the meson build description: https://github.com/mesonbuild/wrapdb/blob/pcre2/subprojects/packagefiles/pcre2/meson.build

In our Github CI, everything works fine on Ubuntu. On Windows using MSVC, I get this error instead:

https://github.com/mesonbuild/wrapdb/runs/4637121959?check_suite_focus=true#step:6:261

Test 8: "Internal offsets and code size tests"

          failed comparison: fc /n D:\a\wrapdb\wrapdb\subprojects\pcre2-10.39\testdata\testoutput8-8-2 testout8\testoutput8-8-2

Do you have any idea what the problem might be? Suggestions for figuring out the problem?

Is this an expected failure? There isn't currently any publicly visible CI for pcre2 (and an open PR only adds some for linux) so it is difficult to know for sure...

[PhilipHazel]

Nobody else has reported problems with Windows, but they may just have ignored this one. I am not a Windows user, and someone else wrote RunTest.bat. Can you post the difference(s) between the two files?

[eli-schwartz]

After some fiddling around with github and entirely too much boilerplate I was able to get it to upload the second file.
testoutput8-8-2.txt

Looks like about 200 lines at the end simply don't exist. :/

I am not a Windows user

Yeah, I understand that feeling...

[eli-schwartz]

lite-xl/lite-xl#785 (comment)

@Jan200101 debugged this and noticed that it reports

Failed: error 184 at offset 1504: (?| and/or (?J: or (?x: parentheses are too deeply nested

[PhilipHazel]

That is not an issue. It is testing that it is detecting that situation (see the output in testoutput-8-8-2). Can this issue now be closed?

[Jan200101]

Yes thats expected, the issue is that stops in the middle of the test (at https://github.com/PhilipHazel/pcre2/blob/31fb2e58a19cfdf5809ff952025b7f4329480337/testdata/testoutput8-8-2#L850 to be exact)

I managed to replicate this on a fresh Windows 10 machine running Visual Studio Community 2022.

Increasing LINK_SIZE from 2 to 4 caused its respective test to pass so I assume its there to be an edge cases specific to Windows

[Jan200101]

Further investigation shows that this was complete nonsense, the issue lies on the wrapdb side.
https://github.com/PhilipHazel/pcre2/blob/31fb2e58a19cfdf5809ff952025b7f4329480337/CMakeLists.txt#L869
This flag is what is missing in the meson.build

[PhilipHazel]

Glad to hear this is resolved. This is not the first time that the small default size of the stack in Windows has been an issue. Sigh. Can this issue now be closed?

[Jan200101]

I believe so

[eli-schwartz]

Yes, I believe so.