netdev - Re: [PATCH v2] usbnet: enable more aggressive autosuspend

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:	Fri, 07 Nov 2008 12:22:36 -0800 (PST)
From:	David Miller <davem@...emloft.net>
To:	oliver@...kum.org
Cc:	jeff@...zik.org, per.hallsmark@...ata.se,
	linux-usb@...r.kernel.org, netdev@...r.kernel.org,
	david-b@...bell.net
Subject: Re: [PATCH v2] usbnet: enable more aggressive autosuspend

From: Oliver Neukum <oliver@...kum.org>
Date: Fri, 7 Nov 2008 13:24:17 +0100

> Am Freitag, 7. November 2008 13:00:05 schrieb Jeff Garzik:
> > Oliver Neukum wrote:
> > > Am Freitag, 7. November 2008 09:26:33 schrieb Jeff Garzik:
> > >> I pondered taking the easy route of fixing this by surrounding the 
> > >> auto_pm reference with an ifdef, but it seems like usbnet could use a 
> > >> bit more thought -- it is questionable whether usbnet_suspend/resume 
> > >> should be built at all, if !CONFIG_PM, even though they are exported.
> > > 
> > > As this is a generic problem, shouldn't we get the compiler to do
> > > this for us?
> > 
> > Can you be more specific?
> 
> As these methods are static the compiler is able to tell whether they
> are referenced by anything but the tables. We should be able to set
> an attribute in the header file that tells the compiler that these methods
> won't be called and can be omitted in the build. Otherwise we have to ifdef
> all those methods.

The problem is that the content of these functions still needs to be
parsed, so references to ifdef'd out structure members are still going
to throw errors.

For the time being please add the necessary CONFIG_PM wrappers around
the suspend and resume methods, as this is what we do tree wide and
I don't think you want these usbnet changes blocked by some fancy
compiler facility that hasn't been implemented yet.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html