[Bug c/57482] New: --help=optimizers reports a wrong list

[Bug c/57482] New: --help=optimizers reports a wrong list

[christophe.beausoleil at sogeti dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57482

            Bug ID: 57482
           Summary: --help=optimizers reports a wrong list
           Product: gcc
           Version: 4.7.3
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: christophe.beausoleil at sogeti dot com

Hello,

I am using GCC 4.7.3 built for AVR targets on a Linux machine. The command :
avr-gcc -mmcu=atmega128 -O1 -Q --help=optimizers
returns a list of enabled options for the -O1 optimization level.

Among these enabled options is :
 -fshort-enums                         [enabled]

However, I can see that this option is not enabled by -O1. I simply did :
typedef enum {val1, val2} Tenum;
Tenum Venum;

I compile with :
avr-gcc -mmcu=atmega128 -O1 test.c -o test.elf

and check Venum size with :
avr-nm -S test.elf | grep Venum

which gives :
00800109 00000002 B Venum

I have to explicitly use -fshort-enums in my command line in order that GCC
takes it into account :
00800109 00000001 B Venum

I must also add that the problem is the same for -00 and -02.

The question is : how can I know which options are effectively enabled or not ?
Can I easily find the answer in source code ?

Thanks a lot for reading till here and for any support !

Best regards

Christophe

[Bug other/57482] [4.7.3][AVR] --help=optimizers reports a wrong list

[gjl at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57482

Georg-Johann Lay <gjl at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|AVR ATmega128               |avr
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2013-06-04
                 CC|                            |gjl at gcc dot gnu.org
          Component|c                           |other
     Ever confirmed|0                           |1
              Build|4.7.3                       |

--- Comment #2 from Georg-Johann Lay <gjl at gcc dot gnu.org> ---
I can confirm this with 4.7.2 and the following source file

== foo.c ==

typedef enum { A, B } ab_t;

int a[sizeof (ab_t) == 2 ? 1 : -1];

== compile ==

$ avr-gcc foo.c

$ avr-gcc foo.c -Q --help=optimizers | grep short.enum
  -fshort-enums                   [enabled]

== issues ==

1) foo.c compiles fine, thus enums are 2 byte per default.
   This shows the output of --help=optimizers is not correct.

2) -f[no-]short-enums is not an optimization option;
   it's an option affecting the ABI.  Better / worse code
   is just a side effect of respective ABI change.

3) The backend does not implement TARGET_DEFAULT_SHORT_ENUMS
   thus the default shall be like with  -fno-short-enums, cf.
   target.def defines the hook default to hook_bool_void_false.

[Bug other/57482] [4.7.3][AVR] --help=optimizers reports a wrong list

[christophe.beausoleil at sogeti dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57482

--- Comment #3 from Christophe <christophe.beausoleil at sogeti dot com> ---
> 2) -f[no-]short-enums is not an optimization option;

Hum, I do not really agree although it is strongly related to ABI, no doubt.
Anyway, it is a very special option as I can see in opts.c :
  /* Set this to a special "uninitialized" value.  The actual default
     is set after target options have been processed.  */
  opts->x_flag_short_enums = 2;

A few specific options are also treated here, but it seems to me that
-fshort-enum is the only optimization option concerned. Am I wrong ? Could it
be the only option which is not reported correctly by --help=optimizers ?

Reading target.def is really instructive, but I still do not understand (yet)
how the optimizations list is built, and how options are overwritten... All
this is very confusing.