This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: use correct DWARF register numbers on rs6000 non-ELF ports



On 17/10/2006, at 7:24 PM, David Edelsohn wrote:


This patch apparently will not change the register numbers for
debugging formats other than DWARF2, but doesn't the move to rs6000.h add
a no-op function call to the debugging path for non-DWARF2 systems?
Slowing down the compiler on AIX isn't appreciated.

Do you think this is a significant effect? For STABS, values from this routine end up in DBX_FINISH_STABS, which makes at least 6 function calls, some of which are to dbxout_int, which calls putc one character at a time to output the value.


<http://gcc.gnu.org/bugzilla/show_bug.cgi?id=3920> indicates that actually, the current behaviour is wrong, and that the function call should not be a no-op for STABS. I didn't fix that because I have no idea what the AIX debugger is expecting.

Attachment: smime.p7s
Description: S/MIME cryptographic signature


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]