[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <4888C232.4090906@sgi.com>
Date: Thu, 24 Jul 2008 10:56:02 -0700
From: Mike Travis <travis@....com>
To: Rusty Russell <rusty@...tcorp.com.au>
CC: Ingo Molnar <mingo@...e.hu>,
Andrew Morton <akpm@...ux-foundation.org>,
"H. Peter Anvin" <hpa@...or.com>, Jack Steiner <steiner@....com>,
Christoph Lameter <cl@...ux-foundation.org>,
linux-kernel@...r.kernel.org, Len Brown <len.brown@...el.com>,
Lennert Buytenhek <kernel@...tstofly.org>,
Dave Jones <davej@...emonkey.org.uk>,
Paul Jackson <pj@....com>, Tony Luck <tony.luck@...el.com>,
Tigran Aivazian <tigran@...azian.fsnet.co.uk>,
Paul Mackerras <paulus@...ba.org>,
Benjamin Herrenschmidt <benh@...nel.crashing.org>,
Robert Richter <robert.richter@....com>,
Martin Schwidefsky <schwidefsky@...ibm.com>,
Heiko Carstens <heiko.carstens@...ibm.com>,
Sam Creasey <sammy@...my.net>, Greg Banks <gnb@....com>,
"Eric W. Biederman" <ebiederm@...ssion.com>,
Adrian Bunk <bunk@...nel.org>,
Thomas Gleixner <tglx@...utronix.de>,
Andreas Schwab <schwab@...e.de>,
Johannes Weiner <hannes@...urebad.de>
Subject: Re: [PATCH 1/1] cpumask: Change cpumask_of_cpu to use cpumask_of_cpu_map
Rusty Russell wrote:
> On Thursday 24 July 2008 03:18:42 Mike Travis wrote:
>> Note that the declaration of cpumask_of_cpu_map[] is initialized
>> so that cpumask_of_cpu(0) is defined early. This assumes that
>> cpu 0 is the boot cpu.
>
> Hi Mike,
>
> Make this statically initialized please. That almost guarantees there'll
> be no problems. It's a little tricky to do, but possible. Patch below
> tested on 32 bit x86 only.
I thought about it, but it didn't seem to be worth the effort. One problem
though, the cpumask bits are such that the LSB of the last word is cpu 0.
So your initializer sets it up in reverse order. I'll see if I can't
figure out how to invert it (very tricky coding btw... ;-)
There was another version of cpumask_of_cpu() that used a percpu variable
for each cpumask_t but I'm not sure where that went. It still required a run
time initializer though (except that cpumask_of_cpu(0) was available early.)
static const cpumask_t cpumask_map[] = {
{ .bits = { 1UL << (0) } }, { .bits = { 1UL << (1) } }, { .bits = { 1UL << (2) } }, { .bits = { 1UL << (3) } },
{ .bits = { 1UL << (4) } }, { .bits = { 1UL << (5) } }, { .bits = { 1UL << (6) } }, { .bits = { 1UL << (7) } },
{ .bits = { 1UL << (8) } }, { .bits = { 1UL << (9) } }, { .bits = { 1UL << (10) } }, { .bits = { 1UL << (11) } },
{ .bits = { 1UL << (12) } }, { .bits = { 1UL << (13) } }, { .bits = { 1UL << (14) } }, { .bits = { 1UL << (15) } },
{ .bits = { 1UL << (16) } }, { .bits = { 1UL << (17) } }, { .bits = { 1UL << (18) } }, { .bits = { 1UL << (19) } },
{ .bits = { 1UL << (20) } }, { .bits = { 1UL << (21) } }, { .bits = { 1UL << (22) } }, { .bits = { 1UL << (23) } },
{ .bits = { 1UL << (24) } }, { .bits = { 1UL << (25) } }, { .bits = { 1UL << (26) } }, { .bits = { 1UL << (27) } },
{ .bits = { 1UL << (28) } }, { .bits = { 1UL << (29) } }, { .bits = { 1UL << (30) } }, { .bits = { 1UL << (31) } },
# 533 ".../linux-2.6.tip/kernel/cpu.c"
{ .bits = { 1UL << (32) } }, { .bits = { 1UL << (33) } }, { .bits = { 1UL << (34) } }, { .bits = { 1UL << (35) } },
{ .bits = { 1UL << (36) } }, { .bits = { 1UL << (37) } }, { .bits = { 1UL << (38) } }, { .bits = { 1UL << (39) } },
{ .bits = { 1UL << (40) } }, { .bits = { 1UL << (41) } }, { .bits = { 1UL << (42) } }, { .bits = { 1UL << (43) } },
{ .bits = { 1UL << (44) } }, { .bits = { 1UL << (45) } }, { .bits = { 1UL << (46) } }, { .bits = { 1UL << (47) } },
{ .bits = { 1UL << (48) } }, { .bits = { 1UL << (49) } }, { .bits = { 1UL << (50) } }, { .bits = { 1UL << (51) } },
{ .bits = { 1UL << (52) } }, { .bits = { 1UL << (53) } }, { .bits = { 1UL << (54) } }, { .bits = { 1UL << (55) } },
{ .bits = { 1UL << (56) } }, { .bits = { 1UL << (57) } }, { .bits = { 1UL << (58) } }, { .bits = { 1UL << (59) } },
{ .bits = { 1UL << (60) } }, { .bits = { 1UL << (61) } }, { .bits = { 1UL << (62) } }, { .bits = { 1UL << (63) } },
>
>> Based on linux-2.6.tip/master at the following commit:
>>
>> commit a099a9b18ab434594bb01ed920e8c58574203fa9
>> Merge: 9b4ae4d... f3b51d7...
>> Author: Ingo Molnar <mingo@...e.hu>
>> Date: Tue Jul 22 15:43:36 2008 +0200
>>
>> Merge branch 'out-of-tree'
>
> This is two-steps forward one step back. Have Ingo drop the original
> patches, and roll together into a single change. That should be much
> clearer. I know it's a PITA.
>
> Thanks,
> Rusty.
>
> Make cpumask_of_cpu_map generic
>
> If an arch doesn't define cpumask_of_cpu_map, create a generic
> statically-initialized one for them. This allows removal of the buggy
> cpumask_of_cpu() macro (&cpumask_of_cpu() gives address of
> out-of-scope var).
>
> An arch with NR_CPUS of 4096 probably wants to allocate this itself
> based on the actual number of CPUs, since otherwise they're using 2MB
> of rodata (1024 cpus means 128k). That's what
> CONFIG_HAVE_CPUMASK_OF_CPU_MAP is for (only x86/64 does so at the
> moment).
>
> In future as we support more CPUs, we'll need to resort to a
> get_cpu_map()/put_cpu_map() allocation scheme.
>
> Signed-off-by: Rusty Russell <rusty@...tcorp.com.au>
>
> diff -r 95e128369c3c include/linux/cpumask.h
> --- a/include/linux/cpumask.h Thu Jul 24 11:16:27 2008 +1000
> +++ b/include/linux/cpumask.h Thu Jul 24 12:33:45 2008 +1000
> @@ -226,23 +226,8 @@ int __next_cpu(int n, const cpumask_t *s
> #define next_cpu(n, src) ({ (void)(src); 1; })
> #endif
>
> -#ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP
> -extern cpumask_t *cpumask_of_cpu_map;
> +extern const cpumask_t *cpumask_of_cpu_map;
> #define cpumask_of_cpu(cpu) (cpumask_of_cpu_map[cpu])
> -
> -#else
> -#define cpumask_of_cpu(cpu) \
> -(*({ \
> - typeof(_unused_cpumask_arg_) m; \
> - if (sizeof(m) == sizeof(unsigned long)) { \
> - m.bits[0] = 1UL<<(cpu); \
> - } else { \
> - cpus_clear(m); \
> - cpu_set((cpu), m); \
> - } \
> - &m; \
> -}))
> -#endif
>
> #define CPU_MASK_LAST_WORD BITMAP_LAST_WORD_MASK(NR_CPUS)
>
> diff -r 95e128369c3c kernel/cpu.c
> --- a/kernel/cpu.c Thu Jul 24 11:16:27 2008 +1000
> +++ b/kernel/cpu.c Thu Jul 24 12:33:45 2008 +1000
> @@ -428,3 +428,112 @@ out:
> #endif /* CONFIG_PM_SLEEP_SMP */
>
> #endif /* CONFIG_SMP */
> +
> +#ifndef CONFIG_HAVE_CPUMASK_OF_CPU_MAP
> +/* 64 bits of zeros, for initializers. */
> +#if BITS_PER_LONG == 32
> +#define Z64 0, 0
> +#else
> +#define Z64 0
> +#endif
> +
> +/* Initializer macros. */
> +#define CMI0(n) { .bits = { 1UL << (n) } }
> +#define CMI(n, ...) { .bits = { __VA_ARGS__, 1UL << ((n) % BITS_PER_LONG) } }
> +
> +#define CMI8(n, ...) \
> + CMI((n), __VA_ARGS__), CMI((n)+1, __VA_ARGS__), \
> + CMI((n)+2, __VA_ARGS__), CMI((n)+3, __VA_ARGS__), \
> + CMI((n)+4, __VA_ARGS__), CMI((n)+5, __VA_ARGS__), \
> + CMI((n)+6, __VA_ARGS__), CMI((n)+7, __VA_ARGS__)
> +
> +#if BITS_PER_LONG == 32
> +#define CMI64(n, ...) \
> + CMI8((n), __VA_ARGS__), CMI8((n)+8, __VA_ARGS__), \
> + CMI8((n)+16, __VA_ARGS__), CMI8((n)+24, __VA_ARGS__), \
> + CMI8((n)+32, 0, __VA_ARGS__), CMI8((n)+40, 0, __VA_ARGS__), \
> + CMI8((n)+48, 0, __VA_ARGS__), CMI8((n)+56, 0, __VA_ARGS__)
> +#else
> +#define CMI64(n, ...) \
> + CMI8((n), __VA_ARGS__), CMI8((n)+8, __VA_ARGS__), \
> + CMI8((n)+16, __VA_ARGS__), CMI8((n)+24, __VA_ARGS__), \
> + CMI8((n)+32, __VA_ARGS__), CMI8((n)+40, __VA_ARGS__), \
> + CMI8((n)+48, __VA_ARGS__), CMI8((n)+56, __VA_ARGS__)
> +#endif
> +
> +#define CMI256(n, ...) \
> + CMI64((n), __VA_ARGS__), CMI64((n)+64, Z64, __VA_ARGS__), \
> + CMI64((n)+128, Z64, Z64, __VA_ARGS__), \
> + CMI64((n)+192, Z64, Z64, Z64, __VA_ARGS__)
> +#define Z256 Z64, Z64, Z64, Z64
> +
> +#define CMI1024(n, ...) \
> + CMI256((n), __VA_ARGS__), \
> + CMI256((n)+256, Z256, __VA_ARGS__), \
> + CMI256((n)+512, Z256, Z256, __VA_ARGS__), \
> + CMI256((n)+768, Z256, Z256, Z256, __VA_ARGS__)
> +#define Z1024 Z256, Z256, Z256, Z256
> +
> +/* We want this statically initialized, just to be safe. We try not
> + * to waste too much space, either. */
> +static const cpumask_t cpumask_map[] = {
> + CMI0(0), CMI0(1), CMI0(2), CMI0(3),
> +#if NR_CPUS > 4
> + CMI0(4), CMI0(5), CMI0(6), CMI0(7),
> +#endif
> +#if NR_CPUS > 8
> + CMI0(8), CMI0(9), CMI0(10), CMI0(11),
> + CMI0(12), CMI0(13), CMI0(14), CMI0(15),
> +#endif
> +#if NR_CPUS > 16
> + CMI0(16), CMI0(17), CMI0(18), CMI0(19),
> + CMI0(20), CMI0(21), CMI0(22), CMI0(23),
> + CMI0(24), CMI0(25), CMI0(26), CMI0(27),
> + CMI0(28), CMI0(29), CMI0(30), CMI0(31),
> +#endif
> +#if NR_CPUS > 32
> +#if BITS_PER_LONG == 32
> + CMI(32, 0), CMI(33, 0), CMI(34, 0), CMI(35, 0),
> + CMI(36, 0), CMI(37, 0), CMI(38, 0), CMI(39, 0),
> + CMI(40, 0), CMI(41, 0), CMI(42, 0), CMI(43, 0),
> + CMI(44, 0), CMI(45, 0), CMI(46, 0), CMI(47, 0),
> + CMI(48, 0), CMI(49, 0), CMI(50, 0), CMI(51, 0),
> + CMI(52, 0), CMI(53, 0), CMI(54, 0), CMI(55, 0),
> + CMI(56, 0), CMI(57, 0), CMI(58, 0), CMI(59, 0),
> + CMI(60, 0), CMI(61, 0), CMI(62, 0), CMI(63, 0),
> +#else
> + CMI0(32), CMI0(33), CMI0(34), CMI0(35),
> + CMI0(36), CMI0(37), CMI0(38), CMI0(39),
> + CMI0(40), CMI0(41), CMI0(42), CMI0(43),
> + CMI0(44), CMI0(45), CMI0(46), CMI0(47),
> + CMI0(48), CMI0(49), CMI0(50), CMI0(51),
> + CMI0(52), CMI0(53), CMI0(54), CMI0(55),
> + CMI0(56), CMI0(57), CMI0(58), CMI0(59),
> + CMI0(60), CMI0(61), CMI0(62), CMI0(63),
> +#endif /* BITS_PER_LONG == 64 */
> +#endif
> +#if NR_CPUS > 64
> + CMI64(64, Z64),
> +#endif
> +#if NR_CPUS > 128
> + CMI64(128, Z64, Z64), CMI64(192, Z64, Z64, Z64),
> +#endif
> +#if NR_CPUS > 256
> + CMI256(256, Z256),
> +#endif
> +#if NR_CPUS > 512
> + CMI256(512, Z256, Z256), CMI256(768, Z256, Z256, Z256),
> +#endif
> +#if NR_CPUS > 1024
> + CMI1024(1024, Z1024),
> +#endif
> +#if NR_CPUS > 2048
> + CMI1024(2048, Z1024, Z1024), CMI1024(3072, Z1024, Z1024, Z1024),
> +#endif
> +#if NR_CPUS > 4096
> +#error NR_CPUS too big. Fix initializers or set CONFIG_HAVE_CPUMASK_OF_CPU_MAP
> +#endif
> +};
> +
> +const cpumask_t *cpumask_of_cpu_map = cpumask_map;
> +#endif /* !CONFIG_HAVE_CPUMASK_OF_CPU_MAP */
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists