[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Tinycc-devel] master bump to mob
From: |
Thomas Preud'homme |
Subject: |
Re: [Tinycc-devel] master bump to mob |
Date: |
Tue, 12 Jun 2012 17:24:20 +0200 |
User-agent: |
KMail/1.13.7 (Linux/3.2.0-2-amd64; KDE/4.7.4; x86_64; ; ) |
Le mardi 12 juin 2012 15:56:57, vous avez écrit :
> Thomas Preud'homme wrote:
> > Greetings grischka,
> >
> > sorry to bother you again but would you mind bumping master to the
> > current mob in the next days? By the way, when do you plan to do a
> > release of tcc?
> >
> > Best regards,
> >
> > Thomas Preud'homme
>
> Done. Btw. you can do that yourself in future if you send me your
> repo.or.cz username so I can give you "master permisssions" for tinycc.
>
> --- grischka
>
> Btw. in
> http://repo.or.cz/w/tinycc.git/commitdiff/9a81dcab0ab1c992efa4a6d74eeb8681
> 2758de20
>
> Is this missing a "break;" or a comment like /* fall through */ ?
>
> case R_ARM_ABS32:
> *(int *)ptr += val;
> break;
> + case R_ARM_REL32:
> + *(int *)ptr += val - addr;
> case R_ARM_BASE_PREL:
> *(int *)ptr += s1->got->sh_addr - addr;
Oh dear, it misses a break indeed. I don't remember the exact test I did but
that was a home made assembly file with no code so I couldn't test the output.
I did however checked the result of the relocation and I remember the
relocation gave a different result on gcc and tcc but that is expected because
the symbol doesn't end up at the same adress in the final executable. So I
tried to compare the computation done in both case and compare with this
formula. Obviously I failed somewhere.
I'll add this break but before committing I'll try to create a full program
which I will be able to execute. Good you noticed.
Are you sure you're still ok with "master permission"? ;-)
I've just created an account robotux on repo.or.cz.
Best regards,
Thomas Preud'homme
signature.asc
Description: This is a digitally signed message part.