linux-kernel - Re: [PATCH 5/5] irqtime: drop local_irq_save/restore from irqtime_account

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <1466695453.15275.31.camel@redhat.com>
Date:	Thu, 23 Jun 2016 11:24:13 -0400
From:	Rik van Riel <riel@...hat.com>
To:	Paolo Bonzini <pbonzini@...hat.com>,
	Peter Zijlstra <peterz@...radead.org>
Cc:	linux-kernel@...r.kernel.org, mingo@...nel.org, fweisbec@...il.com,
	wanpeng.li@...mail.com, efault@....de, tglx@...utronix.de,
	rkrcmar@...hat.com
Subject: Re: [PATCH 5/5] irqtime: drop local_irq_save/restore from
 irqtime_account_irq

On Thu, 2016-06-23 at 15:52 +0200, Paolo Bonzini wrote:
> 
> On 22/06/2016 23:55, Rik van Riel wrote:
> > 
> > > 
> > > > 
> > > > +		hardirq_time =
> > > > READ_ONCE(per_cpu(cpu_hardirq_time,
> > > > cpu));
> > > Which makes this per_cpu(,cpu) usage somewhat curious. What's
> > > wrong
> > > with __this_cpu_read() ?
> > I played around with it a bit, and it seems that
> > __this_cpu_read does not want to nest inside
> > READ_ONCE.  Nobody else seems to be doing that,
> > either.
> According to arch/x86/include/asm/percpu.h, this_cpu_read always has 
> READ_ONCE semantics, but I cannot find that in include/asm-generic
> /percpu.h.  It probably just works because of all the layers of goo,
> but
> something like this (101% untested) would make me feel safer:
> 
Are READ_ONCE semantics desired for every
per-cpu read?

> > Back to READ_ONCE(per_cpu(,cpu)) it is...
> What about READ_ONCE(this_cpu_ptr())?

I tried that yesterday, because it looked like
it would work. Unfortunately, it does not.

  CC      kernel/sched/cputime.o
kernel/sched/cputime.c: In function ‘irqtime_account_irq’:
kernel/sched/cputime.c:107:71: warning: initialization makes pointer
from integer without a cast [-Wint-conversion]
kernel/sched/cputime.c:107:298: error: invalid type argument of unary
‘*’ (have ‘u64 {aka long long unsigned int}’)
kernel/sched/cputime.c:107:428: warning: initialization makes pointer
from integer without a cast [-Wint-conversion]
kernel/sched/cputime.c:107:655: error: invalid type argument of unary
‘*’ (have ‘u64 {aka long long unsigned int}’)
kernel/sched/cputime.c:107:749: warning: initialization makes pointer
from integer without a cast [-Wint-conversion]
kernel/sched/cputime.c:107:976: error: invalid type argument of unary
‘*’ (have ‘u64 {aka long long unsigned int}’)
kernel/sched/cputime.c:107:1087: warning: initialization makes pointer
from integer without a cast [-Wint-conversion]
kernel/sched/cputime.c:107:1314: error: invalid type argument of unary
‘*’ (have ‘u64 {aka long long unsigned int}’)
kernel/sched/cputime.c:107:1408: warning: initialization makes pointer
from integer without a cast [-Wint-conversion]
kernel/sched/cputime.c:107:1635: error: invalid type argument of unary
‘*’ (have ‘u64 {aka long long unsigned int}’)
kernel/sched/cputime.c: In function ‘irqtime_account_hi_update’:
kernel/sched/cputime.c:145:175: warning: comparison of distinct pointer
types lacks a cast
kernel/sched/cputime.c: In function ‘irqtime_account_si_update’:
kernel/sched/cputime.c:162:182: warning: comparison of distinct pointer
types lacks a cast
scripts/Makefile.build:291: recipe for target 'kernel/sched/cputime.o'
failed
make[1]: *** [kernel/sched/cputime.o] Error 1
Makefile:1594: recipe for target 'kernel/sched/' failed
make: *** [kernel/sched/] Error 2

-- 
All rights reversed

Download attachment "signature.asc" of type "application/pgp-signature" (474 bytes)