[PATCH 05/11] omap3: Remove non-existent config option

[Marathe, Yogesh]

> -----Original Message-----
> From: Felipe Contreras [mailto:felipe.contreras at gmail.com]
> Sent: Thursday, September 30, 2010 12:42 AM
> To: Marathe, Yogesh
> Cc: Kanigeri, Hari; Premi, Sanjeev; Tony Lindgren; linux-arm-
> kernel at lists.infradead.org; linux-omap at vger.kernel.org
> Subject: Re: [PATCH 05/11] omap3: Remove non-existent config
> option
> 
> On Wed, Sep 29, 2010 at 4:28 PM, Marathe, Yogesh
> <yogesh_marathe at ti.com> wrote:
> > dsplink and syslink (two drivers who use iommu) should not enable
> > CONFIG_MPU_BRIDGE_IOMMU as dspbridge and dsplink /syslink
> > can not co-exist as they are using same resources. Not applying
> patch
> > breaks dsplink/sylink any one which is being used. Defining this
> config
> > makes them co-exist.
> 
> No, for dsp-link you would have:
> CONFIG_TIDSPBRIDGE=n
> CONFIG_OMAP_IOMMU_IVA2=y
> 
> It would be exactly the same as applying your patch.
> 
> And tidspbridge is not using iommu right now.

I noticed that you have added OMAP_IOMMU_IVA2 to Kconfig. In this
case I need CONFIG_OMAP_IOMMU_IVA2=y by default on master so
that iommu is open to use for all other drivers by default.

> And AFAIK syslink is not for omap3, so omap3_devices is not relevant.
> 
> > I'm ok with changing name to CONFIG_OMAP_IOMMU_IVA2 but
> ideally
> > then that will also break the dspbridge.
> 
> No, grep for MPU_BRIDGE_IOMMU on the current tidspbridge in
> mainline;
> it's not defined anywhere, so CONFIG_OMAP_IOMMU_IVA2, or
> CONFIG_FOOBAR, it doesn't matter for tidspbridge right now. And
> MPU_BRIDGE_IOMMU doesn't depend on tidspbridge on any way.

Please explain, how removing CONFIG_MPU_BRIDGE_IOMMU
Or any other the config name in place, breaks tidspbridge? 
My patch is removing the 'if defined'.

> 
> > One more way would be to soure revert the patch and apply on
> dspbridge branch if it breaks the builds on that branch rather than
> > breaking others in master.
> 
> There is no tidspbrige branch; it's in mainline:
> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-
> 2.6.git;a=tree;f=drivers/staging/tidspbridge
> 
> But that doesn't matter, even if it was in a branch, iommu should not
> break either tidspbridge, or dsp-link, and driver branches should not
> modify anything outside their domain (ideally).
> 
> All you need to do is 'select OMAP_IOMMU_IVA2', although the
> attached
> patch would be needed.
> 
> --
> Felipe Contreras