This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: egcs-980214
- To: Peter Dyballa <pete at riese dot thi dot informatik dot uni-frankfurt dot de>
- Subject: Re: egcs-980214
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Fri, 13 Mar 1998 13:43:48 -0700
- cc: egcs-bugs at cygnus dot com
- Reply-To: law at cygnus dot com
In message <9803131333.AA05113@riese.informatik.uni-frankfurt.de>you write:
> When making everything via a bootstrap-lean it seems that no flags out of
> these
> were allowed "-fomit-frame-pointer -fexpensive-optimizations
> -ffunction-sections -mcpu v8 -m32". I supplied them by defining XCFLAGS or
> TCFLAGS or both when invoking gmake ESC-x compile in GNU Emacs. With
> (almost)
> any of these switches (I did not test each flags exactly and save the huge
> *compilation* buffers) some programmes or libraries are simply not made.
Use BOOT_CFLAGS. Those flags will be used by the stage2/stage3 builds.
Note that on a sparc, -O2 implies -fomit-frame-pointer and
-fexpensive-optimizations. Also nte that -ffunction-sections isn't going
to improve runtime performance without using linker scripts to drive
placement of functions.
> run, always fatal errors. They are due to missing sym links or badly
> installed
You need a recent dejagnu snapshot. They can be found on the egcs
ftp server in the "infrastructure" directory.
> FAIL: gcc.c-torture/execute/loop-2f.c execution, -O2 -g
loop-2f is known to fail on sparcs.
> === g++ Summary ===
>
> # of expected passes 3125
> # of unexpected failures 552
> # of expected failures 79
> # of untested testcases 6
> ===========================================================================
> ==
> + exec ..../egcs-980214/Objekte/../expect/expect -- ..../egcs-980214/gcc/..
> /dejagnu/runtest.exp --tool g77
I would suspect you've got an out of date dejagnu.