C Compilation Issue -- Appears Related to Quoted Items Inside Strings

[bfisherqsi]

There's a hard-to-replicate issue with V that results in bad C code being generated. Compiler version V 0.4.8 f3d0c53 on M2 MAX under MacOS X Sonoma. It seems to be related to a construct like this:

struct MyParams {
     p_fpga_ver  string  @[long: fp_ver; name: 'FPGA Version'; xdoc:'String to use as simulated FPGA version in Version responses. Must be in the form "a.bb.cccc"']
     p_cm_ver    string  @[long: cm_ver; name: 'CM Version'; xdoc:'String to use as simulated CM version in Version responses. Must be in the form "a.bb.cccc"']
}     

Note the quoted elements inside the string, "a.bb.cccc". This structure is being used by the "flag" module.

This issue comes and goes even with the code being unchanged; it's not clear why that's happening, but it happens even when I use -no-parallel. When the error DOES occur, the report points to one of the lines of code related to this structure, showing multiple double-quotes in a row, which is clearly not right. It complains about a parenthesis missing.

Unfortunately, I cannot provide the original code because it's under NDA. And when I tried to make a replication case, it went away.

The workaround is simple: either escape the quotes, or just remove them entirely. But that's just a kludge. I will continue to attempt to isolate a replication case that can be posted. So far, no luck.

Note

You can use the 👍 reaction to increase the issue's priority for developers.

Please note that only the 👍 reaction to the issue itself counts as a vote.
Other reactions and those to comments will not be taken into account.

[medvednikov]

Thank you for reporting this.

[felipensp]

Can you try again with a fresh V version?

[bfisherqsi]

I can try with a newer version when one is pushed, but that's no guarantee, because the bug cannot be reliably replicated even with the original data. My testing was done with version 0.4.8, which is already the most recent release. I updated the version just before running my tests.

[medvednikov]

I'll release 0.4.9 tomorrow.

[felipensp]

I can try with a newer version when one is pushed, but that's no guarantee, because the bug cannot be reliably replicated even with the original data. My testing was done with version 0.4.8, which is already the most recent release. I updated the version just before running my tests.

I meant the v up 👍🏻

[felipensp]

Any news @bfisherqsi ?

[bfisherqsi]

I checked yesterday (‘v up’). The compiler version has not yet changed. Continuing to monitor it.