[Bug target/87508] unrecognized command line option '-mfloat-gprs=double'

[zwilcox at sandia dot gov]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87508

--- Comment #2 from Zachary Wilcox  ---
Thank you sir.  I should have checked the changes.

[Bug target/87508] New: unrecognized command line option '-mfloat-gprs=double'

[zwilcox at sandia dot gov]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87508

Bug ID: 87508
   Summary: unrecognized command line option '-mfloat-gprs=double'
   Product: gcc
   Version: 8.2.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: target
  Assignee: unassigned at gcc dot gnu.org
  Reporter: zwilcox at sandia dot gov
  Target Milestone: ---

The following -mfloat-gprs=double gives an error saying it is an unrecognized
command.

 ./powerpc-eabi-gcc -mfloat-gprs=double
powerpc-sandia_8_1_3-eabi-gcc: error: unrecognized command line option
'-mfloat-gprs=double'; did you mean '-ffloat-store'?
powerpc-sandia_8_1_3-eabi-gcc: fatal error: no input files
compilation terminated.


Even though it is still listed in the documentation:
https://gcc.gnu.org/onlinedocs/gcc/PowerPC-SPE-Options.html

It is not listed  --target-help even though other PowerPC SPE Options are.

I've tried in both GCC 8.2 and 8.1.
It does work in GCC 7.x, 5.x and 4.x.

[Bug debug/87484] internal compiler error: in dwf_regno, at dwarf2cfi.c:909

[zwilcox at sandia dot gov]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87484

Zachary Wilcox  changed:

   What|Removed |Added

 Status|WAITING |RESOLVED
 Resolution|--- |FIXED

--- Comment #2 from Zachary Wilcox  ---
Thank you.
I've moved to 5.4 and the problem has gone away.

Thanks again.

Zach

[Bug debug/87484] New: internal compiler error: in dwf_regno, at dwarf2cfi.c:909

[zwilcox at sandia dot gov]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87484

Bug ID: 87484
   Summary: internal compiler error: in dwf_regno, at
dwarf2cfi.c:909
   Product: gcc
   Version: 4.9.0
Status: UNCONFIRMED
  Severity: normal
  Priority: P3
 Component: debug
  Assignee: unassigned at gcc dot gnu.org
  Reporter: zwilcox at sandia dot gov
  Target Milestone: ---

Created attachment 44774
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=44774=edit
Files that produce errors

Host: Win10
Target: PowerPC
Flags: -O0 -g3 -Wall -mspe=yes -mabi=spe -mfloat-gprs=double -c -mbig-endian
-fno-builtin -fomit-frame-pointer -Wno-unknown-pragmas -mcpu=8548 -MMD -MP -MF

Command line that triggered bug: powerpc-eabi-gcc -O0 -g3 -Wall -mspe=yes
-mabi=spe -mfloat-gprs=double -c -mbig-endian -fno-builtin -fomit-frame-pointer
-Wno-unknown-pragmas -mcpu=8548 -MMD -MP -MF"src/CompilerError.d"
-MT"src/CompilerError.o" -o "src/CompilerError.o" "../src/CompilerError.c"


Compiler output:
 In function 'foo':
../src/CompilerError.c:13:20: warning: unused variable 'lower'
[-Wunused-variable]
  volatile uint64_t lower =  *(ADDR1);
^
../src/CompilerError.c:16:1: internal compiler error: in dwf_regno, at
dwarf2cfi.c:909

Attached is a tar file.
The tar file contains three versions of the code; one that produces the error,
and two that do not, and the preprocessed file