diff options
| author | Simon Marchi <simon.marchi@ericsson.com> | 2015-09-30 11:21:52 -0400 |
|---|---|---|
| committer | Simon Marchi <simon.marchi@ericsson.com> | 2015-09-30 11:21:52 -0400 |
| commit | ede5f15146ae45f4c017f5701629a4fa04ef2beb (patch) | |
| tree | 12bbc83f570aa38aefb58047e9ff74cf76381688 /gdb/gdbarch.h | |
| parent | 7781c06f6e735e92fcf597c578540c2f925a9f05 (diff) | |
| download | gdb-ede5f15146ae45f4c017f5701629a4fa04ef2beb.zip gdb-ede5f15146ae45f4c017f5701629a4fa04ef2beb.tar.gz gdb-ede5f15146ae45f4c017f5701629a4fa04ef2beb.tar.bz2 | |
gdbarch.h: Change gdbarch_info::tdep_info's type to void *
As reported by Ulrich here:
https://sourceware.org/ml/gdb-patches/2015-09/msg00604.html
The system compiler (gcc 4.1) in Centos 5 doesn't like that we cast to a
pointer to a type that doesn't exist. I see no real value in using this
kind iof construct over just using void *. So this patch changes the
tdep_info field to void * and removes the casts. Even in C++, we
should not need an explicit cast when assigning to a void *.
gdb/ChangeLog:
* gdbarch.sh (struct gdbarch_info): Change tdep_info's type to void *.
* gdbarch.h: Regenerate.
* i386-tdep.c (i386_gdbarch_init): Remove cast to
struct gdbarch_tdep_info *.
* mips-tdep.c (mips_gdbarch_init): Likewise.
* ppc-linux-tdep (ppu2spu_sniffer): Likewise.
* rs6000-tdep.c (rs6000_gdbarch_init): Likewise.
* spu-multiarch.c (spu_gdbarch): Likewise.
Diffstat (limited to 'gdb/gdbarch.h')
| -rw-r--r-- | gdb/gdbarch.h | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/gdb/gdbarch.h b/gdb/gdbarch.h index 82e0259..2e4ed3e 100644 --- a/gdb/gdbarch.h +++ b/gdb/gdbarch.h @@ -1561,7 +1561,7 @@ struct gdbarch_info bfd *abfd; /* Use default: NULL (ZERO). */ - struct gdbarch_tdep_info *tdep_info; + void *tdep_info; /* Use default: GDB_OSABI_UNINITIALIZED (-1). */ enum gdb_osabi osabi; |