[PATCH 0/8] low level initialisation for ar2313

antonynpavlov at gmail.com antonynpavlov at gmail.com
Fri Jun 21 03:34:37 EDT 2013


On Fri, 21 Jun 2013 09:23:19 +0200
Sascha Hauer <s.hauer at pengutronix.de> wrote:

> On Fri, Jun 21, 2013 at 10:25:25AM +0400, antonynpavlov at gmail.com wrote:
> > On Fri, 21 Jun 2013 07:30:17 +0200
> > Sascha Hauer <s.hauer at pengutronix.de> wrote:
> > 
> > > Hi Oleksij, Antony,
> > > 
> > > I'm fine with this series. Antony, do you have comments?
> > 
> > The series looks good!
> > 
> > But in the last minute I made a small fix for the "MIPS: pbl: add pbl_sleep macro" patch.
> > 
> > It's a shame, but generated label inside assembly macros are not "local" (at least I have not found how to make them local in my toolchain). 
> > 
> > As the user of the pbl_sleep can use generated label "1" (this label already used in pbl_sleep), there can be some painful side effects, e.g. 
> > 
> > 1:                    // user's generated label
> > 
> >    ...                // some user code
> > 
> >    pbl_sleep t1 100
> > 
> >    ...                // another user code
> > 
> >    b 1b               // jump to the back label "1"
> > 
> > 
> > Inside the pbl_sleep macro we have one more "1" label, so the jump instruction of the
> > example transfer control not to the user's label "1" but to label "1" inside pbl_sleep. 
> > 
> > So I propose a ad-hoc solution.
> > 
> > --- a/arch/mips/include/asm/pbl_macros.h
> > +++ b/arch/mips/include/asm/pbl_macros.h
> > @@ -32,8 +32,8 @@
> >         .set push
> >         .set noreorder
> >         li      \reg, \count
> > -1:
> > -       bgtz    \reg, 1b
> > +254:
> > +       bgtz    \reg, 254b
> >          addi   \reg, -1
> >         .set    pop
> >         .endm
> 
> Yeah, I noticed the same problem some time ago and came up with the same
> solution. I'm not aware of a better way to do this.
> 
> > 
> > The programmers are lazy (at least many of them), they like short names "1", "2"
> > and "254" is very long name, so the conflict is not expected.
> > To prevent conflicts like this we can make a convention "don't use generated labels 200-255 in you assembler pbl code". 
> > 
> > Is there a need for series/patch reposting for fixing the issue?
> 
> Yes, please post what you want me to apply on -next.

Oleksij have reposted the fixed patch: "[PATCH 2/8 v2] MIPS: pbl: add pbl_sleep macro"

> 
> Sascha
> 
> -- 
> Pengutronix e.K.                           |                             |
> Industrial Linux Solutions                 | http://www.pengutronix.de/  |
> Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |


-- 
-- 
Best regards,
  Antony Pavlov



More information about the barebox mailing list