ppcfix.diff

ppcfix.diff

[hugang]

Hi Andrew Morton:

Please apply this patch, to fix compile error in debian woody.
$gcc -v
Reading specs from /usr/lib/gcc-lib/powerpc-linux/2.95.4/specs
gcc version 2.95.4 20011002 (Debian prerelease)

--- 2.6.9-lzf/arch/ppc/syslib/open_pic.c	2004-11-26 12:32:58.000000000 +0800
+++ 2.6.9/arch/ppc/syslib/open_pic.c	2004-11-28 23:16:58.000000000 +0800
@@ -776,7 +776,8 @@ static void openpic_mapirq(u_int irq, cp
 	if (ISR[irq] == 0)
 		return;
 	if (!cpus_empty(keepmask)) {
-		cpumask_t irqdest = { .bits[0] = openpic_read(&ISR[irq]->Destination) };
+		cpumask_t irqdest;
+		irqdest.bits[0] = openpic_read(&ISR[irq]->Destination);
 		cpus_and(irqdest, irqdest, keepmask);
 		cpus_or(physmask, physmask, irqdest);
 	}
-- 
--
Hu Gang / Steve
Linux Registered User 204016
GPG Public Key: http://soulinfo.com/~hugang/hugang.asc

Re: ppcfix.diff

[Jan Engelhardt]

>Please apply this patch, to fix compile error in debian woody.
>$gcc -v
>Reading specs from /usr/lib/gcc-lib/powerpc-linux/2.95.4/specs
>gcc version 2.95.4 20011002 (Debian prerelease)

Is 2.95.*4* supported, after all?



Jan Engelhardt
-- 
Gesellschaft für Wissenschaftliche Datenverarbeitung
Am Fassberg, 37077 Göttingen, www.gwdg.de

Re: ppcfix.diff

[Adrian Bunk]

On Mon, Nov 29, 2004 at 05:42:59PM +0100, Jan Engelhardt wrote:
> >Please apply this patch, to fix compile error in debian woody.
> >$gcc -v
> >Reading specs from /usr/lib/gcc-lib/powerpc-linux/2.95.4/specs
> >gcc version 2.95.4 20011002 (Debian prerelease)
> 
> Is 2.95.*4* supported, after all?

2.95.4 is the version the post-2.95.3 snapshot of the gcc 2.95 
identifies itself. There was no official 2.95.4 and I don't know what is 
supported on ppc, but at least on i386, it is supported.

> Jan Engelhardt

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: ppcfix.diff

[Al Viro]

On Mon, Nov 29, 2004 at 11:40:41PM +0800, hugang@soulinfo.com wrote:
>  	if (!cpus_empty(keepmask)) {
> -		cpumask_t irqdest = { .bits[0] = openpic_read(&ISR[irq]->Destination) };
> +		cpumask_t irqdest;
> +		irqdest.bits[0] = openpic_read(&ISR[irq]->Destination);

Not an equivalent replacement.  The former means "set irqdest.bits[0] to
<expression> and set the rest of fields/array elements in them to zero/null".
The latter leaves everything except irqdest.bits[0] undefined.

Short version of the story: out of
a)	cpumask_t irqdest = { .bits[0] = foo() };
b)	cpumask_t irqdest = { .bits = {[0] = foo()} };
c)	cpumask_t irqdest;
	irqdest.bits[0] = foo();
we have
        valid C89  OK for 2.95  OK for 3.x  valid C99  initializes everything
(a)     no         no           yes         yes        yes
(b)     no         yes          yes         yes        yes
(c)     yes        yes          yes         yes        no

IOW, 2.95 implements only a precursor to C99 initializer syntax and that's
why (a) gives an error.  Proper fix is (b) - replace the line in question
with
             cpumask_t irqdest = { .bits = {[0] = openpic_read(&ISR[irq]->Destination)} };

Re: ppcfix.diff

[Jan Engelhardt]

>>  	if (!cpus_empty(keepmask)) {
>> -		cpumask_t irqdest = { .bits[0] = openpic_read(&ISR[irq]->Destination) };
>> +		cpumask_t irqdest;
>> +		irqdest.bits[0] = openpic_read(&ISR[irq]->Destination);
>
>Not an equivalent replacement.  The former means "set irqdest.bits[0] to
><expression> and set the rest of fields/array elements in them to zero/null".

Really? I thought zero'ing only happens for static storage.



Jan Engelhardt
-- 
Gesellschaft für Wissenschaftliche Datenverarbeitung
Am Fassberg, 37077 Göttingen, www.gwdg.de

Re: ppcfix.diff

[Al Viro]

On Mon, Nov 29, 2004 at 06:44:14PM +0100, Jan Engelhardt wrote:
> >>  	if (!cpus_empty(keepmask)) {
> >> -		cpumask_t irqdest = { .bits[0] = openpic_read(&ISR[irq]->Destination) };
> >> +		cpumask_t irqdest;
> >> +		irqdest.bits[0] = openpic_read(&ISR[irq]->Destination);
> >
> >Not an equivalent replacement.  The former means "set irqdest.bits[0] to
> ><expression> and set the rest of fields/array elements in them to zero/null".
> 
> Really? I thought zero'ing only happens for static storage.

Two different issues - behaviour when no initializer is given and behaviour
when initializer does not cover everything.

IOW,

static int a1[4];		/* all zeroes, since it's non-auto */
static int a2[4] = {1,0,0,0}	/* 1 0 0 0, we have initialized them all */
static int a3[4] = {1}		/* 1 0 0 0, missing members are zeroed */
void f(void)
{
	int b1[4];		/* undefined contents */
	int b2[4] = {1,0,0,0}	/* 1 0 0 0, we have initialized them all */
	int b3[4] = {1}		/* 1 0 0 0, missing members are zeroed */

C99 6.7.8(21) and 6.7.8(10) deal with "not covers everything" and
"no initializer at all" resp.