linux-kernel - Re: [PATCH] dmaengine: fsl-edma: fix Makefile logic

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, 11 Jan 2024 11:09:22 -0500
From: Frank Li <Frank.li@....com>
To: Arnd Bergmann <arnd@...db.de>
Cc: Arnd Bergmann <arnd@...nel.org>, Vinod Koul <vkoul@...nel.org>,
	Peng Fan <peng.fan@....com>, Fabio Estevam <festevam@...x.de>,
	dmaengine@...r.kernel.org, linux-kernel@...r.kernel.org,
	imx@...ts.linux.dev
Subject: Re: [PATCH] dmaengine: fsl-edma: fix Makefile logic

On Thu, Jan 11, 2024 at 07:23:34AM +0100, Arnd Bergmann wrote:
> On Thu, Jan 11, 2024, at 01:02, Frank Li wrote:
> > On Thu, Jan 11, 2024 at 12:03:42AM +0100, Arnd Bergmann wrote:
> >> From: Arnd Bergmann <arnd@...db.de>
> >> 
> >> A change to remove some unnecessary exports ended up removing some
> >> necessary ones as well, and caused a build regression by trying to
> >> link a single source file into two separate modules:
> >
> > You should fix Kconfig to provent fsl-edma and mcf-edma build at the same
> > time.
> 
> That sounds like the wrong approach since it prevents
> compile-testing one of the drivers in an allmodconfig
> build.
> 
> > EXPORT_SYMBOL_GPL is not necesary at all.
> >
> > mcf-edma is quit old. ideally, it should be merged into fsl-edma.
> 
> I have no specific interest in either of the drivers, just
> trying to fix the build regression. I see no harm in exporting
> the symbols, but I can refactor the drivers to link all three
> files into the same module and add a hack to register both
> platform_driver instances from a shared module_init() if
> you think that's better. Unfortunately we can't have more
> than one initcall in a loadable module.

It should be better link into same module because some debugfs and trace
improvement are on my TODO list. Export symbols will make more unnessary
complex. Or simple exclude MCF_EDMA by change Kconfig

config MCF_EDMA                                                            
        tristate "Freescale eDMA engine support, ColdFire mcf5441x SoCs"   
        depends on !FSL_EDMA
		   ^^^^^
	depends on M5441x || COMPILE_TEST 


> 
>      Arnd