This is the mail archive of the gdb-patches@sources.redhat.com mailing list for the GDB 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: [commit/multiarch/hppa64] further pa64 multiarching changes...


Hello,

I have committed the following change which brings hppa64-*-hpux11 to
the same level of multiarching as the 32bit hppa.

Some comments:
- I have elected to take this opportunity to create hppa-tdep.h,
although it was not absolutely necessary for now. Since I think
I will eventually need it, I just went ahead now. In retrospect
(already? :-), I am not sure this is such a good reason...

Actually, the sooner the better. Gets it over and done with :-)
Seriously, for some reason people are reluctant to add new .h files even going as far as using extern's in .c files. Creating the file up front gets that over and done with.


  - The patch is a bit large, as it contains several changes that could
    have been broken down into smaller patches. I am sorry about that.
    My only excuse is that I am currently working in degraded mode,
    since Ada Core is still in the dark.

  - Instead of duplicating functions adding 64bit counterparts, I
    elected to use the gdbarch_tdep structure to hold the address
    size (in number of bytes), and use this value in some of the methods
    to determine how to compute the result. Comments welcome.

It's a judgment call. The MIPS, which now has separate functions, has too many different ABIs to contend with - fix one, break another - hence the split. Doesn't sound like you've got that problem.


Andrew

(Things to do is re-structure MIPS push arguments so that it is part shared and part ABI specific)



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