#698 closed defect (invalid)
Tcl/tk won't build properly with gcc 3.3.2
Reported by: | Owned by: | ||
---|---|---|---|
Priority: | lowest | Milestone: | |
Component: | Book | Version: | CVS |
Severity: | normal | Keywords: | |
Cc: |
Description
In stage 1 if you build tcl with gcc 3.3.2,(with the host gcc Fedora .94, 3.2.2) you get an odd error message, it is probally an incompatibility with gcc 3.3.2 or a bug in the new gcc, ether way I thought I report it. (note: it does compile correctly with gcc 3.3.1 on my system)
CVS Oct 23/2003 Fedora .94 HP pavilion 533c(pentium 4, 512mb ram, few gigabytes reserved for each GNU Linux partion) build partion reiserfs(latest at time) buildlt both corectly and incorretly with Glibc CVS (yea, can you believe it)
post and additional questions as comments
Attachments (1)
Change History (4)
by , 21 years ago
comment:1 by , 21 years ago
Resolution: | → invalid |
---|---|
Status: | new → closed |
This looks to be an invalid bug report. The file it's choking on (compat/strstr.c) shouldn't even get built. This suggests something went astray during your LFS build. Look at the tcl configure log, it should say this:
checking proper strstr implementation... yes
If not, then the steps in the book were not followed exactly.
By the time you compile TCL, you are pretty much divorced from the host system which also suggests something went wrong for you. I'd suggest restarting from scratch and be sure to perform all sanity checks along the way.
Cannot reproduce here with gcc-3.3.2, Closing.
comment:2 by , 21 years ago
Let's see if it is a bug. 1) Google for ISO-C-FDIS.1999-04.pdf, download it 2) Look in Appendix B for headers that define NULL. They are: <locale.n> <sttdef.h> <stdio.h> <stdlib.h> <string.h> <time.h> <wchar.h> The file in question (compat/strstr.c) doesn't include anything of the above. So it's a bug in Tcl, but it should never trigger on LFS for the reasons outlined by Greg.
comment:3 by , 21 years ago
I was talking about when you install tcl for the toolchain test, but you were right I have recently discovered something went wrong when I compiled Glibc(stage 1), currently I'm almost baffled by it.
log