Hi Arnd, On Tue, May 3, 2011 at 3:18 PM, Arnd Bergmann <arnd@xxxxxxxx> wrote: > On Thursday 28 April 2011 17:31:46 Omar Ramirez Luna wrote: >> Rework and rebase of the previous set of patches for iommu migration[1]. >> >> Reorganized the patches to avoid several iterations to change the same code. >> >> Patches were tested with an OMAP3630 board, with both, basic samples and >> gst-dsp; these are dependant on patch: >> "OMAP: iovmm: fix SW flags passed by user"[2]. > > The code removal in this series looks good, but the move from > one proprietary iommu API to another proprietary iommu API > seems pointless. > > It would be much more helpful if you could convert the tidspbridge > code to the dma-mapping API, or to the generic iommu API > and provide these interfaces from the omap code. Yes, I thought removing a bunch of lines from staging might be good. I'll check what can be done to use generic iommu or dma-mapping. Thanks, Omar _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
