linux-kernel - Re: [PATCH 0/5] Convert to use devm

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:   Sun, 26 Mar 2017 23:38:55 +0800
From:   Leo Yan <leo.yan@...aro.org>
To:     Alexandre Belloni <alexandre.belloni@...e-electrons.com>
Cc:     Russell King <linux@...linux.org.uk>,
        Dmitry Torokhov <dmitry.torokhov@...il.com>,
        Alessandro Zummo <a.zummo@...ertech.it>,
        Linus Walleij <linus.walleij@...aro.org>,
        Baptiste Reynal <b.reynal@...tualopensystems.com>,
        Alex Williamson <alex.williamson@...hat.com>,
        Jaroslav Kysela <perex@...ex.cz>,
        Takashi Iwai <tiwai@...e.com>, linux-input@...r.kernel.org,
        linux-kernel@...r.kernel.org, rtc-linux@...glegroups.com,
        linux-arm-kernel@...ts.infradead.org, kvm@...r.kernel.org,
        alsa-devel@...a-project.org
Subject: Re: [PATCH 0/5] Convert to use devm_*() for amba attached modules

On Sun, Mar 26, 2017 at 05:20:50PM +0200, Alexandre Belloni wrote:
> On 26/03/2017 at 22:41:49 +0800, Leo Yan wrote:
> > When review device driver modules which attach to amba bus, found
> > several modules are not using devm_*() apis to manage resource. As
> > result, some drivers have memory leakage or missing iomem unmapping
> > when rmmod module. And the code has many "goto" tags to handle
> > different failures.
> > 
> > So this patch series is to convert to use devm_*() for moudules which
> > are attached to amba bus to manage resource and get more robust and
> > neat code.
> > 
> > Patch 0003 "drivers/rtc/rtc-pl031.c: Convert to use devm_*()" has been
> > verified on 96boards Hikey. Other patches can pass building but have
> > not really tested on hardware.
> > 
> 
> If your plan is to actually remove usage of
> amba_request_regions() and amba_release_regions(), you should do so in
> its own patch sets instead of hiding that in a useless cleanup series.

Just curious, from Russell's replying for patch 0005, IIUC we cannot
totally remove usage of amba_request_regions() and
amba_release_regions(), there have some coner case should use
amba_request_regions() + ioremap().

Does it make sense to remove most usage of amba_request_regions() and
amba_release_regions() but we still keep these two functions in the kernel?

Thanks,
Leo Yan