Török Edwin wrote:
[snip]
>>> I built 0.94rc1 myself on cygwin, and confirmed this.
>>>
>>> For some reason lstat() works with 0.94 ... after we dropped cygwin support.
>>> Weird isn't it?
>>>
>> No, programming is deterministic, something changed in the code.
>>
>
> My guess is that one of the C_CYGWIN portions of code actually made
> things worse, and caused clamdscan to fail.
> But nobody noticed, until now, when we added unit tests.
Found it, I told you it was non-sense:
In file clamdscan/client.c at line 251, as you said, there's a C_CYGWIN
section... which to me makes no sense, so I commented it out,
//#ifdef C_CYGWIN
// sprintf(fullpath, "%s", filename);
//#else
if(!getcwd(cwd, PATH_MAX)) {
logg("^Can't get absolute pathname of current working
directory.\n");
return NULL;
}
sprintf(fullpath, "%s/%s", cwd, filename);
//#endif
Now, re-testing (clamd is 0.94 but it doesn't matter, I tested before
and after the change, before fails, after works... OK, I'll show both
tests).
$ ../clamdscan/clamdscan -h
ClamAV Daemon Client 0.94rc1
...
(I used -h since -V shows the daemon version -- more weirdness)
$ ../clamdscan/clamdscan clam.exe
clam.exe: lstat() failed. ERROR
$ vi client.c
... commented out the lines ...
$ cd ..
$ make
... (only client.c was recompiled and clamdscan relinked)
$ ../clamdscan/clamdscan clam.exe
/home/rberber/Projects/clamav-0.94rc1/test/clam.exe: ClamAV-Test-File FOUND
----------- SCAN SUMMARY -----------
Infected files: 1
Time: 0.139 sec (0 m 0 s)
Do we get back support for Cygwin?
--
René Berber
_______________________________________________
Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net
http://www.clamav.net/support/ml
