linux-kernel - Re: [PATCH v2 1/2] Staging: comedi: convert while loops to timeouts in s626.c

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]

Date:	Thu, 6 Mar 2014 00:13:50 -0800 (PST)
From:	Chase Southwood <chase.southwood@...oo.com>
To:	Ian Abbott <abbotti@....co.uk>,
	"gregkh@...uxfoundation.org" <gregkh@...uxfoundation.org>
Cc:	Ian Abbott <ian.abbott@....co.uk>,
	"hsweeten@...ionengravers.com" <hsweeten@...ionengravers.com>,
	"devel@...verdev.osuosl.org" <devel@...verdev.osuosl.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 1/2] Staging: comedi: convert while loops to timeouts in s626.c

>On Wednesday, March 5, 2014 6:10 AM, Ian Abbott <abbotti@....co.uk> wrote:

>>On 2014-03-04 08:43, Chase Southwood wrote:
>>This patch changes a handful of while loops to timeouts to prevent
>>infinite looping on hardware failure.  A couple such loops are in a
>>function (s626_debi_transfer()) which is called from critical sections,
>>so comedi_timeout() is unusable for them, and an iterative timeout is
>>used instead.  For the while loops in a context where comedi_timeout() is
>>allowed, a new callback function, s626_send_dac_eoc(), has been defined
>>to evaluate the conditions that the while loops are testing.  The new
>>callback employs a switch statement based on the register offset values
>>that the status is to be checked from, as this information is enough to
>>recreate the entire readl() call, and is passed in the 'context'
>>parameter.  A couple of additional macros have been defined where this
>>method is not completely sufficient.  The proper comedi_timeout() calls
>>are then used.
>>
>>Signed-off-by: Chase Southwood <chase.southwood@...oo.com>
>>---
>>Ian, my idea for the callback function here is a bit iffy, I'd love to
>>hear your feedback.  I've created the ret variable because patch 2/2
>>propogates all errors upwards.
>>
>>2: Now using comedi_timeout() where allowable.

[snip]

>Rather than switching on a value that is either a register offset or 
>some special value, it might be better to use an enum to define the 
>context values you are switching on, e.g.:
>
>enum {
>    s626_send_dac_wait_not_mc1_a2out,
>    s626_send_dac_wait_ssr_af2_out,
>    s626_send_dac_wait_fb_buffer2_msb_00,
>    s626_send_dac_wait_fb_buffer2_msb_ff
>};
>
>Alternatively, different callback functions could be used for each case.

Ian,
Ah yes...this is much smarter than what I was trying before.  I've made
up a patch using this enum, and tomorrow I will get PATCH 2/2 rebased
and get a v2 of this patchset sent out.

Thanks,
Chase

>One slight oddity about the original code that waits for the MSB of FB 
>BUFFER2 to go from 0x00 to 0xFF is that it actually waits for it to 
>become non-zero.  I'm presuming the intermediate values between 0x00 and 
>0xFF are never seen.  (There is a non-Comedi, GPL'ed, Linux driver 
>available from Sensoray that makes the same assumption, but I think the 
>Comedi driver was derived from it.)
>

[snip]
>-- 
>-=( Ian Abbott @ MEV Ltd.    E-mail: <abbotti@....co.uk>        )=-
>-=( Tel: +44 (0)161 477 1898   FAX: +44 (0)161 718 3587         )=-
>-- 
--
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/