On Sat, 19 Feb 2022 16:53:46 -0700, Brian Inglis
> On 2022-02-19 16:37, Lemures Lemniscati wrote:
> > On Sat, 19 Feb 2022 10:11:06 -0700, Brian Inglis
> >> On 2022-02-18 22:26, Lemures Lemniscati wrote:
> >>> On Fri, 18 Feb 2022 12:33:30 -0700, Brian Inglis
> >>>> Any idea why autoconf tests for iconv now fail with undefined refs to 
> >>>> iconv{_open,,_close}! >>>> STC attached, log, -E output. Tried also with 
> >>>> -l iconv.
> 
> >>> https://cygwin.com/cgi-bin2/jobs.cgi?srcpkg=playground&id=3861
> >>> This will successfully build libisocodes-1.2.4 but some tests fail.
> 
> >> Not so worried about that, just noticed that iconv conftest also failed in 
> >> my recent findutils build which I released, missing libiconv functions and 
> >> presumably substituting gnulib functions, although the iconv conftest 
> >> passed in github scallywag.
> >> This may be the result of a change in gnulib or autoconf, which now 
> >> sometimes fails when it should not.
> 
> > Although `gcc iconv-test.c -liconv` should pass, it might have been used
> > for test, a command like either `gcc -liconv iconv-test.c`,
> > or `gcc iconv-test.c`.
> 
> After conftest if iconv... is in libc, the next conftest compiles adding 
> -liconv, and still fails.
> I'm going to look for most recent successful iconv conftest in config.log and 
> compare the logs and configure scripts iconv sections to those in the failing 
> findutils and libisocodes, to see if I can track it back to gnulib or 
> autoconf changes to the compile flags or options, or at least enough info to 
> file a bug report upstream somewhere.


It seems ok on my local environment. The iconv part of the log is attached.

```
cygport findutils.cygport download finish prep compile
cat -n findutils-4.9.0-1.x86_64/build/config.log | tail +35145 | head -676 > 
config_L35145.log
```

Lem

Attachment: config_L35145.log
Description: Binary data

-- 
Problem reports:      https://cygwin.com/problems.html
FAQ:                  https://cygwin.com/faq/
Documentation:        https://cygwin.com/docs.html
Unsubscribe info:     https://cygwin.com/ml/#unsubscribe-simple

Reply via email to