gcc problem?
Eric Christopher
echristo@cygnus.com
Thu Oct 5 13:23:00 GMT 2000
Julie,
If you could preprocess the source it would be appreciated. The best
way to do this would be to do the compile with -save-temps and mail the
.i file. (and yes, this is a bug in gcc)
-eric
Julie Mitchell wrote:
>
> Hello,
>
> I have recently been trying to compile LAS (Land Analysis System) on
> a Linux laptop. In compiling a particular function for a library I am
> getting a strange error. Below is the pertinent information, and
> I have attached the source code for the offending function, c_fhdrrd.c.
> Please let me know if you need anything else, and what ideas you may
> have.
>
> Thank you!
>
> Julie Mitchell
> Email: jmitch@kryos.colorado.edu
> ------------------------------------------
>
> GCC version 2.95.2
>
> System: SuSE Linux 6.4 (i386) - Kernel 2.2.14 (2)
>
> Compiler command:
>
> gcc -O -funsigned-char -fwritable-strings -Wall
> -I/usr/local/las73/build/las/include
> -I/usr/local/las73/tae/include -c c_fhdrrd.c
>
> Error generated:
>
> c_fhdrrd.c: In function `c_fhdrrd':
> c_fhdrrd.c:209: Unable to generate reloads for:
> (insn 795 793 797 (parallel[
> (set (reg:SI 0 %eax)
> (fix:SI (fix:DF (subreg:DF (reg:DI 0 %eax) 0))))
> (clobber (mem:HI (plus:SI (reg:SI 6 %ebp)
> (const_int -386 [0xfffffe7e])) 0))
> (clobber (mem:HI (plus:SI (reg:SI 6 %ebp)
> (const_int -388 [0xfffffe7c])) 0))
> (clobber (mem:SI (plus:SI (reg:SI 6 %ebp)
> (const_int -392 [0xfffffe78])) 0))
> (clobber (scratch:HI))
> ] ) 149 {fix_truncdfsi2+1} (insn_list 793 (nil))
> (expr_list:REG_EQUIV (mem:SI (reg/v:SI 4 %esi) 0)
> (expr_list:REG_DEAD (reg:DI 0 %eax)
> (expr_list:REG_UNUSED (scratch:HI)
> (nil)))))
> make[1]: *** [c_fhdrrd.o] Error 1
> make[1]: Leaving directory `/usr/local/las73/src/las/base/support/header'
> make: *** [las] Error 2
> --------------------------------------------
>
> ------------------------------------------------------------------------
> Name: c_fhdrrd.c
> c_fhdrrd.c Type: Plain Text (text/plain)
> Description: Text
More information about the Gcc-bugs
mailing list