Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 27 Sep 2011 18:56:13 +0300
From:      Alexander Motin <mav@FreeBSD.org>
To:        Adrian Chadd <adrian@freebsd.org>
Cc:        freebsd-current@freebsd.org
Subject:   Re: ath / 802.11n performance issues and timer code
Message-ID:  <4E81F21D.7000209@FreeBSD.org>
In-Reply-To: <4E81E4A1.3070301@FreeBSD.org>
References:  <CAJ-VmomZyDJV62yCQOvG=UB6H4wfz9=3_cWzEL7vWAA14TCyYA@mail.gmail.com>	<201109261053.30410.jhb@freebsd.org>	<CAJ-VmomyMQpcpHo4ve-_O67_8VVJW-YRqWeb-7exSxzV5rz_pA@mail.gmail.com>	<201109261305.57602.jhb@freebsd.org>	<CAJ-Vmo=ZBwBu6Mk=XdY1p18s5=kQvmC3qZc1JzPtpppGBs4Z=Q@mail.gmail.com>	<CAJ-Vmong0dJN=t=Qq4%2BaLyBGxEKAKBd8oX4y9P0goux%2BZAc3yA@mail.gmail.com> <CAJ-VmonbAgsNjdCstd_Ap6JBqowD1NX0J5rQ=t9ideaiXTXd%2BA@mail.gmail.com> <4E81E4A1.3070301@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
Alexander Motin wrote:
> Adrian Chadd wrote:
>> .. erm, sys/mips/mips/machdep.c:
>>
>> /*
>>  * call platform specific code to halt (until next interrupt) for the idle loop
>>  */
>> void
>> cpu_idle(int busy)
>> {
>>         KASSERT((mips_rd_status() & MIPS_SR_INT_IE) != 0,
>>                 ("interrupts disabled in idle process."));
>>         KASSERT((mips_rd_status() & MIPS_INT_MASK) != 0,
>>                 ("all interrupts masked in idle process."));
>>
>>         if (!busy) {
>>                 critical_enter();
>>                 cpu_idleclock();
>>         }
>>         __asm __volatile ("wait");
>>         if (!busy) {
>>                 cpu_activeclock();
>>                 critical_exit();
>>         }
>> }
>>
>> .. does that look right?
> 
> Yes it does. x86 does the same, but with more details. The general idea
> of the critical section is to block context switch out of idle thread
> until missed time events will be handled inside cpu_activeclock().

I was wrong. That's not good. I have no idea about mips wait instruction
semantics, related to disabling interrupts. In x86 semantics proper
solution is:

	disable_intr();
	if (sched_runnable())
		enable_intr();
	else
		__asm __volatile("sti; hlt");

It makes interrupts enabled atomically with entering sleep state, that
closes race window and prevents entering into sleep after receiving
interrupt.

-- 
Alexander Motin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4E81F21D.7000209>