linux-kernel - Re: [PATCH] clk: Add tracepoints for hardware operations

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: <53B209E5.1050701@codeaurora.org>
Date:	Mon, 30 Jun 2014 18:07:49 -0700
From:	Stephen Boyd <sboyd@...eaurora.org>
To:	Steven Rostedt <rostedt@...dmis.org>
CC:	Mike Turquette <mturquette@...aro.org>,
	linux-kernel@...r.kernel.org, linux-arm-msm@...r.kernel.org,
	linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH] clk: Add tracepoints for hardware operations

On 06/30/14 17:52, Steven Rostedt wrote:
> On Mon, 30 Jun 2014 16:56:39 -0700
> Stephen Boyd <sboyd@...eaurora.org> wrote:
>
>> @@ -483,10 +486,12 @@ static void clk_unprepare_unused_subtree(struct clk *clk)
>>  		return;
>>  
>>  	if (__clk_is_prepared(clk)) {
>> +		trace_clk_unprepare(clk);
> Does it make sense to do these when clk->ops->unprepared_unused or
> uprepare is not set?
>
> You can use DEFINE_EVENT_CONDITIONAL() and add as condition:
>
>    clk->ops->unprepared_unused || clk->ops->unprepare
>

Neat. I don't know if we actually want to do that though. If we always
record an event even when the hardware doesn't support the operation we
get information about events happening to the clock from a software
perspective. If that isn't important, then we can probably just put it
under the if conditions.

>
>>  		if (clk->ops->enable) {
>>  			ret = clk->ops->enable(clk->hw);
>>  			if (ret) {
>> @@ -945,6 +965,7 @@ static int __clk_enable(struct clk *clk)
>>  				return ret;
> It may make even more sense to add the tracepoints within the if
> statement. Especially if you have a return on error.
>
>

Right. I was thinking that no "clk*_complete" event would mean there was
some error. Detecting that case is not so easy though. It may be better
to always have the completion event so we know how long hardware
operations take and so that error handling is simpler.

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation

--
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/