From: Peter Zijlstra <peterz@infradead.org>
To: Shakeel Butt <shakeel.butt@linux.dev>
Cc: Vlastimil Babka <vbabka@suse.cz>,
Andrew Morton <akpm@linux-foundation.org>,
Tejun Heo <tj@kernel.org>, Johannes Weiner <hannes@cmpxchg.org>,
Michal Hocko <mhocko@kernel.org>,
Roman Gushchin <roman.gushchin@linux.dev>,
Muchun Song <muchun.song@linux.dev>,
Alexei Starovoitov <ast@kernel.org>,
Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
bpf@vger.kernel.org, linux-mm@kvack.org, cgroups@vger.kernel.org,
linux-kernel@vger.kernel.org,
Meta kernel team <kernel-team@meta.com>,
Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
npiggin@gmail.com
Subject: Re: [PATCH 0/4] memcg: nmi-safe kmem charging
Date: Wed, 14 May 2025 09:11:44 +0200 [thread overview]
Message-ID: <20250514071144.GB24938@noisy.programming.kicks-ass.net> (raw)
In-Reply-To: <ct2h2eyuepa2g2ltl5fucfegwyuqspvz6d4uugcs4szxwnggdc@6m4ks3hp3tjj>
On Tue, May 13, 2025 at 03:17:00PM -0700, Shakeel Butt wrote:
> > IIRC Power64 has issues here, 'funnily' their local_t is NMI safe.
> > Perhaps we could do the same for their this_cpu_*(), but ideally someone
> > with actual power hardware should do this ;-)
> >
>
> Is CONFIG_ARCH_HAS_NMI_SAFE_THIS_CPU_OPS the right config to
> differentiate between such archs? I see Power64 does not have that
> enabled.
> > There is no config symbol for this presently.
>
> Hmm what about CONFIG_ARCH_HAS_NMI_SAFE_THIS_CPU_OPS?
Hmm, I didn't know about that one, and it escaped my grep yesterday.
Anyway, PPC is fixable, just not sure its worth it for them.
> >
> > > - (if the above leaves any 64bit arch) its 64bit atomics implementation is safe
> >
> > True, only because HPPA does not in fact have NMIs.
>
> What is HPPA?
arch/parisc the worst 64bit arch ever.
They saw sparc32-smp and thought that was a great idea, or something
along those lines. Both are quite terrible. Sparc64 realized the mistake
and fixed it -- it has cmpxchg.
Nick, is this something that's useful for you guys?
---
diff --git a/arch/powerpc/include/asm/percpu.h b/arch/powerpc/include/asm/percpu.h
index ecf5ac70cfae..aa188db68ef5 100644
--- a/arch/powerpc/include/asm/percpu.h
+++ b/arch/powerpc/include/asm/percpu.h
@@ -25,6 +25,11 @@ DECLARE_STATIC_KEY_FALSE(__percpu_first_chunk_is_paged);
#define percpu_first_chunk_is_paged false
#endif
+#ifdef CONFIG_PPC_BOOK3S_64
+#define __pcpu_local_irq_save(f) powerpc_local_irq_pmu_save(f)
+#define __pcpu_local_irq_restore(s) powerpc_local_irq_pmu_restore(f)
+#endif
+
#include <asm-generic/percpu.h>
#include <asm/paca.h>
diff --git a/include/asm-generic/percpu.h b/include/asm-generic/percpu.h
index 02aeca21479a..5c8376588dfb 100644
--- a/include/asm-generic/percpu.h
+++ b/include/asm-generic/percpu.h
@@ -75,6 +75,11 @@ extern void setup_per_cpu_areas(void);
#define PER_CPU_ATTRIBUTES
#endif
+#ifndef __pcpu_local_irq_save
+#define __pcpu_local_irq_save(x) raw_local_irq_save(x)
+#define __pcpu_local_irq_restore(x) raw_local_irq_restore(x)
+#endif
+
#define raw_cpu_generic_read(pcp) \
({ \
*raw_cpu_ptr(&(pcp)); \
@@ -146,9 +151,9 @@ do { \
({ \
TYPEOF_UNQUAL(pcp) ___ret; \
unsigned long ___flags; \
- raw_local_irq_save(___flags); \
+ __pcpu_local_irq_save(___flags); \
___ret = raw_cpu_generic_read(pcp); \
- raw_local_irq_restore(___flags); \
+ __pcpu_local_irq_restore(___flags); \
___ret; \
})
@@ -165,9 +170,9 @@ do { \
#define this_cpu_generic_to_op(pcp, val, op) \
do { \
unsigned long __flags; \
- raw_local_irq_save(__flags); \
+ __pcpu_local_irq_save(__flags); \
raw_cpu_generic_to_op(pcp, val, op); \
- raw_local_irq_restore(__flags); \
+ __pcpu_local_irq_restore(__flags); \
} while (0)
@@ -175,9 +180,9 @@ do { \
({ \
TYPEOF_UNQUAL(pcp) __ret; \
unsigned long __flags; \
- raw_local_irq_save(__flags); \
+ __pcpu_local_irq_save(__flags); \
__ret = raw_cpu_generic_add_return(pcp, val); \
- raw_local_irq_restore(__flags); \
+ __pcpu_local_irq_restore(__flags); \
__ret; \
})
@@ -185,9 +190,9 @@ do { \
({ \
TYPEOF_UNQUAL(pcp) __ret; \
unsigned long __flags; \
- raw_local_irq_save(__flags); \
+ __pcpu_local_irq_save(__flags); \
__ret = raw_cpu_generic_xchg(pcp, nval); \
- raw_local_irq_restore(__flags); \
+ __pcpu_local_irq_restore(__flags); \
__ret; \
})
@@ -195,9 +200,9 @@ do { \
({ \
bool __ret; \
unsigned long __flags; \
- raw_local_irq_save(__flags); \
+ __pcpu_local_irq_save(__flags); \
__ret = raw_cpu_generic_try_cmpxchg(pcp, ovalp, nval); \
- raw_local_irq_restore(__flags); \
+ __pcpu_local_irq_restore(__flags); \
__ret; \
})
@@ -205,9 +210,9 @@ do { \
({ \
TYPEOF_UNQUAL(pcp) __ret; \
unsigned long __flags; \
- raw_local_irq_save(__flags); \
+ __pcpu_local_irq_save(__flags); \
__ret = raw_cpu_generic_cmpxchg(pcp, oval, nval); \
- raw_local_irq_restore(__flags); \
+ __pcpu_local_irq_restore(__flags); \
__ret; \
})
next prev parent reply other threads:[~2025-05-14 7:11 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-05-09 23:28 Shakeel Butt
2025-05-09 23:28 ` [PATCH 1/4] memcg: add infra for nmi safe memcg stats Shakeel Butt
2025-05-09 23:28 ` [PATCH 2/4] memcg: add nmi-safe update for MEMCG_KMEM Shakeel Butt
2025-05-09 23:28 ` [PATCH 3/4] memcg: nmi-safe slab stats updates Shakeel Butt
2025-05-09 23:28 ` [PATCH 4/4] memcg: make objcg charging nmi safe Shakeel Butt
2025-05-13 22:25 ` Alexei Starovoitov
2025-05-14 16:46 ` Shakeel Butt
2025-05-10 1:26 ` [PATCH 0/4] memcg: nmi-safe kmem charging Andrew Morton
2025-05-10 3:11 ` Shakeel Butt
2025-05-10 7:00 ` Harry Yoo
2025-05-12 14:52 ` Vlastimil Babka
2025-05-12 15:56 ` Vlastimil Babka
2025-05-12 19:12 ` Shakeel Butt
2025-05-13 7:15 ` Vlastimil Babka
2025-05-13 11:41 ` Peter Zijlstra
2025-05-13 22:17 ` Shakeel Butt
2025-05-14 7:11 ` Peter Zijlstra [this message]
2025-05-15 1:49 ` Shakeel Butt
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=20250514071144.GB24938@noisy.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=akpm@linux-foundation.org \
--cc=ast@kernel.org \
--cc=bigeasy@linutronix.de \
--cc=bpf@vger.kernel.org \
--cc=cgroups@vger.kernel.org \
--cc=hannes@cmpxchg.org \
--cc=kernel-team@meta.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mathieu.desnoyers@efficios.com \
--cc=mhocko@kernel.org \
--cc=muchun.song@linux.dev \
--cc=npiggin@gmail.com \
--cc=roman.gushchin@linux.dev \
--cc=shakeel.butt@linux.dev \
--cc=tj@kernel.org \
--cc=vbabka@suse.cz \
/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