Mail Archives: djgpp/2015/07/01/23:50:20
On 2.7.2015 0:57, Juan Manuel Guerrero (juan DOT guerrero AT gmx DOT de) [via
djgpp AT delorie DOT com] wrote:
> Am 30.06.2015 19:47, schrieb Eli Zaretskii (eliz AT gnu DOT org):
>>> Date: Tue, 30 Jun 2015 10:11:44 -0700 (PDT)
>>> From: "Juan Manuel Guerrero (juan DOT guerrero AT gmx DOT de)"<djgpp AT delorie DOT com>
>>>
>>> Can you tell me what binutils you are using? I do not know if it is
>>> of any importance but we can try some older version that bnu22b.
>>
>> I have Binutils 2.19 here.
>
> OFYI, I have recompiled gdb771 using gcc510, bnu225br2 and djdev205.
> Then I have compiled the following test program using gcc346b, bnu225br2
> and djdev205:
>
> -------------------- b.c start --------------------
> -------------------- b.c end --------------------
>
> -------------------- a.c start --------------------
> -------------------- a.c end --------------------
>
> -------------------- gdb.txt start --------------------
> b main
> b 16
> b 7
> b b.c : 6
> r
> info source
> -------------------- gdb.txt end --------------------
>
> I am still not able to step into certain functions. Specially if the
> functions
> are defined in different files. I was able to step from main into print().
> But I was not able to step from inside print() into do_print(). Please
> note
> that the argument of do_print is the function allocate_string() defined
> in b.c.
>
> But if I write a control file for gdb like gdb.txt then it becomes
> possible to
> stop in every function no matter in which file it is defined. This may
> be a
> little bit cumbersome but it seems to work. I know that this is
> certainly not
> a "complex" program but I think that it demonstrates that a combination
> of gcc346b,
> bnu225br2, gdb771b (to be recompiled with djdev205) and djdev205 may be
> usefull
> to continue developing emacs. Of course, in the end it must really be
> tried by
> compiling emacs. Anything else will never give a definitive answer to
> this issue.
I tried also my year 2009 build of gdb-6.8 (never uploaded to
ftp.delorie.com but most likely source of snapshot build Eli is using)
It behaves in exactly same way. So current problems is not a regression
since that time. I do not think that we can fix these problems without
considerable work.
Andris
PS. My old build is still at http://ap1.pp.fi/djgpp/gdb/6.8/ and I gave
this link sometimes to Eli
(http://www.delorie.com/djgpp//mail-archives/browse.cgi?p=djgpp/2009/05/03/03:21:46)
- Raw text -