linux-kernel - RE: [Query] mwifiex: few observations to reduce number of endian conversions

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: <ee79492a163d4b9baef75e28c7a93791@SC-EXCH04.marvell.com>
Date:	Tue, 21 Jun 2016 10:17:42 +0000
From:	Amitkumar Karwar <akarwar@...vell.com>
To:	Prasun Maiti <prasunmaiti87@...il.com>,
	Nishant Sarmukadam <nishants@...vell.com>
CC:	Linux Kernel <linux-kernel@...r.kernel.org>,
	Linux Next <linux-next@...r.kernel.org>,
	WiFi Mailing List <linux-wireless@...r.kernel.org>,
	Johannes Berg <johannes@...solutions.net>
Subject: RE: [Query] mwifiex: few observations to reduce number of endian
 conversions

Hi Prasun,

> From: Prasun Maiti [mailto:prasunmaiti87@...il.com]
> Sent: Friday, June 17, 2016 6:42 PM
> To: Amitkumar Karwar; Nishant Sarmukadam
> Cc: Linux Kernel; Linux Next; WiFi Mailing List; Johannes Berg
> Subject: [Query] mwifiex: few observations to reduce number of endian
> conversions
> 
> Hi Amitkumar,
> 
> I have two observations:
> 
> 1. I have found that in the command response path for host command
> "HostCmd_CMD_802_11_EEPROM_ACCESS", a "0" value has been endian
> converted. It can only be a safe futuristic approach for any non-zero
> value there however! Otherwise, the endian conversion can be removed.
> 
> 2. For multiple Host Commands (e.g HostCmd_CMD_802_11_EEPROM_ACCESS
> etc.) "cpu_to_leX"-converted values are saved to driver. So "leX_to_cpu"
> conversion is required too many times afterwards in driver.
> On the contrary, we can save the values to driver without any
> conversion, and only command buffer(s) are prepared with endian
> converted values. In this way we can gain some efficiency [code size /
> time] by reducing the number of endian conversion considerably.
> 

I agree with your observations. We'll prepare a cleanup patch to address this.

Regards,
Amitkumar