From: Suresh Siddha <suresh.b.siddha@intel.com>
To: Alexander Gordeev <agordeev@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>,
linux-kernel@vger.kernel.org, x86@kernel.org,
Yinghai Lu <yinghai@kernel.org>
Subject: Re: [PATCH 6/8] x86/apic: Avoid useless scanning thru a cpumask in assign_irq_vector()
Date: Thu, 07 Jun 2012 15:28:49 -0700 [thread overview]
Message-ID: <1339108129.28766.109.camel@sbsiddha-desk.sc.intel.com> (raw)
In-Reply-To: <20120607131542.GE4759@dhcp-26-207.brq.redhat.com>
On Thu, 2012-06-07 at 15:15 +0200, Alexander Gordeev wrote:
> In case of static vector allocation domains (i.e. flat) if all vector
> numbers are exhausted, an attempt to assign a new vector will lead to
> useless scans through all CPUs in the cpumask, even though it is known
> that each new pass would fail. Make this corner case less painful by
> letting report whether the vector allocation domain depends on passed
> arguments or not and stop scanning early.
>
> The same could have been achived by introducing a static flag to the
> apic operations. But let's allow vector_allocation_domain() have more
> intelligence here and decide dynamically, in case we would need it in
> the future.
Acked-by: Suresh Siddha <suresh.b.siddha@intel.com>
>
> Signed-off-by: Alexander Gordeev <agordeev@redhat.com>
> ---
> arch/x86/include/asm/apic.h | 8 +++++---
> arch/x86/kernel/apic/apic_noop.c | 3 ++-
> arch/x86/kernel/apic/io_apic.c | 12 +++++++++---
> 3 files changed, 16 insertions(+), 7 deletions(-)
>
> diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h
> index feb2dbd..e3fecd5 100644
> --- a/arch/x86/include/asm/apic.h
> +++ b/arch/x86/include/asm/apic.h
> @@ -306,7 +306,7 @@ struct apic {
> unsigned long (*check_apicid_used)(physid_mask_t *map, int apicid);
> unsigned long (*check_apicid_present)(int apicid);
>
> - void (*vector_allocation_domain)(int cpu, struct cpumask *retmask);
> + bool (*vector_allocation_domain)(int cpu, struct cpumask *retmask);
> void (*init_apic_ldr)(void);
>
> void (*ioapic_phys_id_map)(physid_mask_t *phys_map, physid_mask_t *retmap);
> @@ -615,7 +615,7 @@ extern unsigned int
> default_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
> const struct cpumask *andmask);
>
> -static inline void
> +static inline bool
> flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
> {
> /* Careful. Some cpus do not strictly honor the set of cpus
> @@ -628,12 +628,14 @@ flat_vector_allocation_domain(int cpu, struct cpumask *retmask)
> */
> cpumask_clear(retmask);
> cpumask_bits(retmask)[0] = APIC_ALL_CPUS;
> + return false;
> }
>
> -static inline void
> +static inline bool
> default_vector_allocation_domain(int cpu, struct cpumask *retmask)
> {
> cpumask_copy(retmask, cpumask_of(cpu));
> + return true;
> }
>
> static inline unsigned long default_check_apicid_used(physid_mask_t *map, int apicid)
> diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c
> index 3e43cf5..ac9edf2 100644
> --- a/arch/x86/kernel/apic/apic_noop.c
> +++ b/arch/x86/kernel/apic/apic_noop.c
> @@ -100,11 +100,12 @@ static unsigned long noop_check_apicid_present(int bit)
> return physid_isset(bit, phys_cpu_present_map);
> }
>
> -static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask)
> +static bool noop_vector_allocation_domain(int cpu, struct cpumask *retmask)
> {
> if (cpu != 0)
> pr_warning("APIC: Vector allocated for non-BSP cpu\n");
> cpumask_copy(retmask, cpumask_of(cpu));
> + return true;
> }
>
> static u32 noop_apic_read(u32 reg)
> diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
> index 05af3d3..4061a7d 100644
> --- a/arch/x86/kernel/apic/io_apic.c
> +++ b/arch/x86/kernel/apic/io_apic.c
> @@ -1137,8 +1137,9 @@ __assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask)
> for_each_cpu_and(cpu, mask, cpu_online_mask) {
> int new_cpu;
> int vector, offset;
> + bool more_domains;
>
> - apic->vector_allocation_domain(cpu, tmp_mask);
> + more_domains = apic->vector_allocation_domain(cpu, tmp_mask);
>
> if (cpumask_subset(tmp_mask, cfg->domain)) {
> free_cpumask_var(tmp_mask);
> @@ -1153,8 +1154,13 @@ next:
> offset = (offset + 1) % 16;
> vector = FIRST_EXTERNAL_VECTOR + offset;
> }
> - if (unlikely(current_vector == vector))
> - continue;
> +
> + if (unlikely(current_vector == vector)) {
> + if (more_domains)
> + continue;
> + else
> + break;
> + }
>
> if (test_bit(vector, used_vectors))
> goto next;
> --
> 1.7.7.6
>
next prev parent reply other threads:[~2012-06-07 22:29 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-06-05 11:23 [PATCH 4/8] x86: apic: Factor out default vector_allocation_domain() operations Alexander Gordeev
2012-06-06 8:22 ` Ingo Molnar
2012-06-06 8:42 ` Alexander Gordeev
2012-06-07 13:14 ` [PATCH 4/8] x86/apic: Factor out default vector_allocation_domain() operation Alexander Gordeev
2012-06-07 22:24 ` Suresh Siddha
2012-06-08 14:49 ` [tip:x86/apic] " tip-bot for Alexander Gordeev
2012-06-07 13:15 ` [PATCH 5/8] x86/apic: Try to spread IRQ vectors to different priority levels Alexander Gordeev
2012-06-07 22:26 ` Suresh Siddha
2012-06-08 14:50 ` [tip:x86/apic] " tip-bot for Alexander Gordeev
2012-06-20 17:23 ` H. Peter Anvin
2012-06-20 21:41 ` Suresh Siddha
2012-06-20 21:46 ` H. Peter Anvin
2012-06-07 13:15 ` [PATCH 6/8] x86/apic: Avoid useless scanning thru a cpumask in assign_irq_vector() Alexander Gordeev
2012-06-07 22:28 ` Suresh Siddha [this message]
2012-06-08 14:51 ` [tip:x86/apic] " tip-bot for Alexander Gordeev
2012-06-07 13:15 ` [PATCH 7/8] x86/apic: Make cpu_mask_to_apicid() operations return error code Alexander Gordeev
2012-06-07 22:24 ` Suresh Siddha
2012-06-08 15:15 ` Alexander Gordeev
2012-06-08 16:53 ` [PATCH] x86/apic: Eliminate cpu_mask_to_apicid() operation Alexander Gordeev
2012-06-08 18:24 ` Suresh Siddha
2012-06-08 22:36 ` Yinghai Lu
2012-06-11 8:22 ` Ingo Molnar
2012-06-11 10:51 ` Alexander Gordeev
2012-06-08 14:51 ` [tip:x86/apic] x86/apic: Make cpu_mask_to_apicid() operations return error code tip-bot for Alexander Gordeev
2012-06-07 13:16 ` [PATCH 8/8] x86/apic: Make cpu_mask_to_apicid() operations check cpu_online_mask Alexander Gordeev
2012-06-08 14:52 ` [tip:x86/apic] " tip-bot for Alexander Gordeev
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1339108129.28766.109.camel@sbsiddha-desk.sc.intel.com \
--to=suresh.b.siddha@intel.com \
--cc=agordeev@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=x86@kernel.org \
--cc=yinghai@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®