[Bug 323528] Re: mktime bug back with gcc 4.3.2
bug in the testcase. see gcc(1) -fwrapv how to work around this. ** Changed in: gcc-4.3 (Ubuntu) Status: Confirmed => Invalid -- mktime bug back with gcc 4.3.2 https://bugs.launchpad.net/bugs/323528 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 323528] Re: mktime bug back with gcc 4.3.2
For the record, re-running autoconf on the given project's files (archfs), updated the configure script so that it no longer writes the test in this way. The test then passes appropriately with Ubuntu 8.10. I'll file a bug against the archfs project to update their configure script. -- mktime bug back with gcc 4.3.2 https://bugs.launchpad.net/bugs/323528 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 323528] Re: mktime bug back with gcc 4.3.2
Really, all that's needed is this part, which basically occurs on line 222 of the above bar.c: void main() { int i = 0; for (i = 1; 0 < i; i *= 2) { if (0) return; } return; } As you can see, the loop is designed to end when i wraps around and goes negative. With -O2, this doesn't happen for some reason. -- mktime bug back with gcc 4.3.2 https://bugs.launchpad.net/bugs/323528 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 323528] Re: mktime bug back with gcc 4.3.2
I can confirm this on 32bit fully updated 8.10 as well. After some more testing, I can confirm this is a gcc bug because it is the -O2 flag which causes the problem. Without that, it is fine. I have attached the source code to the test to this bug report. Compile like so: $ gcc -o bar1 -O2 bar.c $ gcc -o bar2 bar.c $ ./bar1 [hangs] $ ./bar2 [no hang] (-O1 also has no problems) ** Attachment added: "bar.c" http://launchpadlibrarian.net/24970275/bar.c ** Changed in: gcc-4.3 (Ubuntu) Status: Incomplete => Confirmed -- mktime bug back with gcc 4.3.2 https://bugs.launchpad.net/bugs/323528 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 323528] Re: mktime bug back with gcc 4.3.2
please attach the source and header files used by the autoconf test ** Changed in: gcc-4.3 (Ubuntu) Status: New => Incomplete -- mktime bug back with gcc 4.3.2 https://bugs.launchpad.net/bugs/323528 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 323528] Re: mktime bug back with gcc 4.3.2
** Changed in: gcc-4.3 (Ubuntu) Sourcepackagename: gcc-defaults => gcc-4.3 -- mktime bug back with gcc 4.3.2 https://bugs.launchpad.net/bugs/323528 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 323528] Re: mktime bug back with gcc 4.3.2
** Changed in: gcc-defaults (Ubuntu) Sourcepackagename: autoconf => gcc-defaults -- mktime bug back with gcc 4.3.2 https://bugs.launchpad.net/bugs/323528 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs