On Wed, 2011-05-18 at 18:27 +0200, Samuel Thibault wrote:
> Svante Signell, le Wed 18 May 2011 17:41:50 +0200, a écrit :
> > On Wed, 2011-05-18 at 17:08 +0200, Samuel Thibault wrote:
> > > Svante Signell, le Wed 18 May 2011 16:54:24 +0200, a écrit :
> > ..
> > > > Is this OK? There is also a similar case in the glibc code.
> > >
> > > Where?
> >
> > See below:
>
> Ok, the same fix applies indeed.
>
> > Changing "1" and "2" to "m" made the warnings disappear. But
> > you say that this is not the correct solution. Care to explain what that
> > code is doing and what the changes mean?
>
> Well, actually in that case it'll be the same. But it's much more clear
> to announce the memory as being both read&written ("+m") than announcing
> it as read and also as written ("m" and "=m").
Can you tell if the change has any effect on the produced code or not?
At least we get rid of a lot of warnings when compiling, and according
to the output no registers are used without the change?? Did you change
that code, and if so how?
> For more details, see tutorials & documentation about assembly
> constraints.
I'm updating myself, both gcc asm and asm seems to be needed :)
