Description of problem: I just tried to compile tla-1.3-1 from Redhat Fedora Extras development tree with the Intel C compiler. The compiler said 1. /usr/src/redhat/BUILD/tla-1.3/src/hackerlab/char/dstr.c(325): remark #592: variable "a" is used before its value is set /usr/src/redhat/BUILD/tla-1.3/src/hackerlab/char/dstr.c(325): remark #592: variable "a_l" is used before its value is set /usr/src/redhat/BUILD/tla-1.3/src/hackerlab/char/dstr.c(325): remark #592: variable "b" is used before its value is set /usr/src/redhat/BUILD/tla-1.3/src/hackerlab/char/dstr.c(325): remark #592: variable "b_l" is used before its value is set The source code is int dstr_cmp (t_dstr ad, t_dstr bd) { t_uchar * a; size_t a_l; t_uchar * b; size_t b_l; return str_cmp_n (a, a_l, b, b_l); } This looks like very wrong code. Is it work in progress ? 2. /usr/src/redhat/BUILD/tla-1.3/src/hackerlab/char/dstr.c(344): remark #592: variable "a" is used before its value is set /usr/src/redhat/BUILD/tla-1.3/src/hackerlab/char/dstr.c(344): remark #592: variable "a_l" is used before its value is set /usr/src/redhat/BUILD/tla-1.3/src/hackerlab/char/dstr.c(344): remark #592: variable "b" is used before its value is set /usr/src/redhat/BUILD/tla-1.3/src/hackerlab/char/dstr.c(344): remark #592: variable "b_l" is used before its value is set Duplicate. 3. /usr/src/redhat/BUILD/tla-1.3/src/tla/libarch/merge-points.c(108): warning #192: unrecognized character escape sequence The source code is safe_printfmt (2, "arch: attempt to compute merge-points of non-existent revision\n (%s\%s)\n", into_arch->name, into_spec); \% is a valid specifier. Was %%%s intended ? Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
1 and 2 definitely look . . . bad. I'll send those upstream. But number 3 is not a bug (as there are 2 '%s' escapes and two strings). Before going any furthur, could you try recompiling tla-1.3.1-4.fix.1 SRPM from the devel repository? There are some other issues in 1.3.1 at present, but yours might be fixed.
>But number 3 is not a bug (as there are 2 '%s' escapes and two strings). I'm sorry, I wrote the wrong thing. What I should have said was that the escape sequence \% isn't a valid escape sequence. I've just checked K&R 2, and page 38 in my copy says it isn't either. Suggest rework, but I am unable to suggest a replacement, since I don't know what is required here. >could you try recompiling tla-1.3.1-4.fix.1 SRPM from >the devel repository? Sorry, no. That machine no longer runs Redhat Linux. I suggest that using gcc compiler flag -Wall might get you what you want.
problems 1 and 2 have been fixed and 3 doesn't cause any errors and pretty much isn't going to be fixed upstream, so I'm closing as WONTFIX.