linux-kernel - Re: per-cpu operation madness vs validation

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

Message-ID: <alpine.DEB.2.00.1107271155410.13058@router.home>
Date:	Wed, 27 Jul 2011 12:04:18 -0500 (CDT)
From:	Christoph Lameter <cl@...ux.com>
To:	Peter Zijlstra <peterz@...radead.org>
cc:	Tejun Heo <tj@...nel.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	"Paul E. McKenney" <paulmck@...ibm.com>,
	linux-kernel <linux-kernel@...r.kernel.org>,
	Ingo Molnar <mingo@...e.hu>, eric.dumazet@...il.com
Subject: Re: per-cpu operation madness vs validation

On Wed, 27 Jul 2011, Peter Zijlstra wrote:

> Also things like the below hunk are just plain ugly and obfuscate the
> code to safe one load at best. I'm sorely tempted to revert such crap.
>
> @@ -1468,14 +1465,12 @@ static void x86_pmu_start_txn(struct pmu *pmu)
>   */
>  static void x86_pmu_cancel_txn(struct pmu *pmu)
>  {
> -       struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
> -
> -       cpuc->group_flag &= ~PERF_EVENT_TXN;
> +       __this_cpu_and(cpu_hw_events.group_flag, ~PERF_EVENT_TXN);
>         /*
>          * Truncate the collected events.
>          */
> -       cpuc->n_added -= cpuc->n_txn;
> -       cpuc->n_events -= cpuc->n_txn;
> +       __this_cpu_sub(cpu_hw_events.n_added, __this_cpu_read(cpu_hw_events.n_txn));
> +       __this_cpu_sub(cpu_hw_events.n_events, __this_cpu_read(cpu_hw_events.n_txn));
>         perf_pmu_enable(pmu);
>  }

Yea that is pretty ugly and went overboard. Too many segment
overrides there. The following patch will make that easier to read. We
could also remove the whole hunk. __this_cpu ops there only causes the
generation of simple xadds avoiding register operations.

---
 arch/x86/kernel/cpu/perf_event.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

Index: linux-2.6/arch/x86/kernel/cpu/perf_event.c
===================================================================
--- linux-2.6.orig/arch/x86/kernel/cpu/perf_event.c	2011-07-27 11:59:35.000000000 -0500
+++ linux-2.6/arch/x86/kernel/cpu/perf_event.c	2011-07-27 12:03:41.000000000 -0500
@@ -1612,12 +1612,14 @@ static void x86_pmu_start_txn(struct pmu
  */
 static void x86_pmu_cancel_txn(struct pmu *pmu)
 {
+	int x = __this_cpu_read(cpu_hw_events.n_txn);
+
 	__this_cpu_and(cpu_hw_events.group_flag, ~PERF_EVENT_TXN);
 	/*
 	 * Truncate the collected events.
 	 */
-	__this_cpu_sub(cpu_hw_events.n_added, __this_cpu_read(cpu_hw_events.n_txn));
-	__this_cpu_sub(cpu_hw_events.n_events, __this_cpu_read(cpu_hw_events.n_txn));
+	__this_cpu_sub(cpu_hw_events.n_added, x);
+	__this_cpu_sub(cpu_hw_events.n_events, x);
 	perf_pmu_enable(pmu);
 }





--
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/