Message ID | 1480600080-30196-1-git-send-email-benjamin.gaignard@linaro.org (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
+ Vladimir 2016-12-01 14:48 GMT+01:00 Benjamin Gaignard <benjamin.gaignard@linaro.org>: > commit ab6494f0c96f ("nommu: Add noMMU support to the DMA API") have > add CONFIG_MMU compilation flag but that prohibit to use dma_mmap_wc() > when the platform doesn't have MMU. > > This patch call vm_iomap_memory() in noMMU case to test if addresses > are correct and set wma->vm_flags rather than all return an error. > > Signed-off-by: Benjamin Gaignard <benjamin.gaignard@linaro.org> > Cc: Catalin Marinas <catalin.marinas@arm.com> > Cc: arnd@arndb.de > --- > arch/arm/mm/dma-mapping.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c > index ab4f745..230875e 100644 > --- a/arch/arm/mm/dma-mapping.c > +++ b/arch/arm/mm/dma-mapping.c > @@ -868,6 +868,9 @@ static int __arm_dma_mmap(struct device *dev, struct vm_area_struct *vma, > vma->vm_end - vma->vm_start, > vma->vm_page_prot); > } > +#else > + ret = vm_iomap_memory(vma, vma->vm_start, > + (vma->vm_end - vma->vm_start)); > #endif /* CONFIG_MMU */ > > return ret; > -- > 1.9.1 >
On 09/12/16 15:27, Benjamin Gaignard wrote: > + Vladimir I'm not in DMA, but I can see that with your patch it deviates from dma_common_mmap(), can you give more context, please? Cheers Vladimir > > 2016-12-01 14:48 GMT+01:00 Benjamin Gaignard <benjamin.gaignard@linaro.org>: >> commit ab6494f0c96f ("nommu: Add noMMU support to the DMA API") have >> add CONFIG_MMU compilation flag but that prohibit to use dma_mmap_wc() >> when the platform doesn't have MMU. >> >> This patch call vm_iomap_memory() in noMMU case to test if addresses >> are correct and set wma->vm_flags rather than all return an error. >> >> Signed-off-by: Benjamin Gaignard <benjamin.gaignard@linaro.org> >> Cc: Catalin Marinas <catalin.marinas@arm.com> >> Cc: arnd@arndb.de >> --- >> arch/arm/mm/dma-mapping.c | 3 +++ >> 1 file changed, 3 insertions(+) >> >> diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c >> index ab4f745..230875e 100644 >> --- a/arch/arm/mm/dma-mapping.c >> +++ b/arch/arm/mm/dma-mapping.c >> @@ -868,6 +868,9 @@ static int __arm_dma_mmap(struct device *dev, struct vm_area_struct *vma, >> vma->vm_end - vma->vm_start, >> vma->vm_page_prot); >> } >> +#else >> + ret = vm_iomap_memory(vma, vma->vm_start, >> + (vma->vm_end - vma->vm_start)); >> #endif /* CONFIG_MMU */ >> >> return ret; >> -- >> 1.9.1 >> >
2016-12-09 16:36 GMT+01:00 Vladimir Murzin <vladimir.murzin@arm.com>: > On 09/12/16 15:27, Benjamin Gaignard wrote: >> + Vladimir > > I'm not in DMA, but I can see that with your patch it deviates from > dma_common_mmap(), can you give more context, please? I'm working on ARM platform with MMU (stm32f4) to enable display driver. Framebuffer is allocated with dma_alloc_wc() and when userland try to mmap drm/kms fraemwork calls dma_mmap_wc(). All this is in drivers/gpu/drm/drm_gem_cma_helper.c dma_mmap_wc() call failed because __arm_dma_mmap() always return an error if CONFIG_MMU isn't defined. That what I try to solve with this patch. > Cheers > Vladimir > >> >> 2016-12-01 14:48 GMT+01:00 Benjamin Gaignard <benjamin.gaignard@linaro.org>: >>> commit ab6494f0c96f ("nommu: Add noMMU support to the DMA API") have >>> add CONFIG_MMU compilation flag but that prohibit to use dma_mmap_wc() >>> when the platform doesn't have MMU. >>> >>> This patch call vm_iomap_memory() in noMMU case to test if addresses >>> are correct and set wma->vm_flags rather than all return an error. >>> >>> Signed-off-by: Benjamin Gaignard <benjamin.gaignard@linaro.org> >>> Cc: Catalin Marinas <catalin.marinas@arm.com> >>> Cc: arnd@arndb.de >>> --- >>> arch/arm/mm/dma-mapping.c | 3 +++ >>> 1 file changed, 3 insertions(+) >>> >>> diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c >>> index ab4f745..230875e 100644 >>> --- a/arch/arm/mm/dma-mapping.c >>> +++ b/arch/arm/mm/dma-mapping.c >>> @@ -868,6 +868,9 @@ static int __arm_dma_mmap(struct device *dev, struct vm_area_struct *vma, >>> vma->vm_end - vma->vm_start, >>> vma->vm_page_prot); >>> } >>> +#else >>> + ret = vm_iomap_memory(vma, vma->vm_start, >>> + (vma->vm_end - vma->vm_start)); >>> #endif /* CONFIG_MMU */ >>> >>> return ret; >>> -- >>> 1.9.1 >>> >> >
On Thu, Dec 01, 2016 at 02:48:00PM +0100, Benjamin Gaignard wrote: > commit ab6494f0c96f ("nommu: Add noMMU support to the DMA API") have > add CONFIG_MMU compilation flag but that prohibit to use dma_mmap_wc() > when the platform doesn't have MMU. > > This patch call vm_iomap_memory() in noMMU case to test if addresses > are correct and set wma->vm_flags rather than all return an error. ITYM vma. I think this is fine, but I've no way to know as I don't run nommu here... I've been hoping that others working on nommu would at least give an ack or something, but I guess not. So, please put it in the patch system, thanks.
diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index ab4f745..230875e 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -868,6 +868,9 @@ static int __arm_dma_mmap(struct device *dev, struct vm_area_struct *vma, vma->vm_end - vma->vm_start, vma->vm_page_prot); } +#else + ret = vm_iomap_memory(vma, vma->vm_start, + (vma->vm_end - vma->vm_start)); #endif /* CONFIG_MMU */ return ret;
commit ab6494f0c96f ("nommu: Add noMMU support to the DMA API") have add CONFIG_MMU compilation flag but that prohibit to use dma_mmap_wc() when the platform doesn't have MMU. This patch call vm_iomap_memory() in noMMU case to test if addresses are correct and set wma->vm_flags rather than all return an error. Signed-off-by: Benjamin Gaignard <benjamin.gaignard@linaro.org> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: arnd@arndb.de --- arch/arm/mm/dma-mapping.c | 3 +++ 1 file changed, 3 insertions(+)