Description of problem: I just had a look at the source code for glibc-2.3.4-3 I found the following assignment statements nested into assert statement. ../BUILD/glibc-20041219T2331/iconv/strtab.c: assert (endp = retval + st->total + 1); ../BUILD/glibc-20041219T2331/stdio-common/tst-ungetc.c: assert ((c = getc (fp)) == 'a'); ../BUILD/glibc-20041219T2331/hurd/sigunwind.c: assert (link->thread.prevp = &ss->active_resources); I'm not saying this code is wrong - just deeply suspicous. The code will do different things, depending on wether the assert statements are switched in or out. Suggest code rework to avoid such assignment statements in assert statements. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
Thanks. strtab.c and sigunwind.c fixed in CVS, tst-ungetc.c doesn't use the normal assert, but defines its own assert macro, so there is no bug there.