Linaro-mm-sig September 2019

linaro-mm-sig@lists.linaro.org

7 participants
6 discussions

by Christian König

Hi everyone, since upstreaming the full dynamic DMA-buf changes turned out more problematic than previously thought I've reverted back to individual patches and separated out only the locking changes. So this patch does NOT contain any new callbacks for pinning/unpinning and move notification, but only the locking changes necessary. As previously discussed when the framework detects that the locking semantics between exporter and importer are different it just falls back to using a cached sgt created during attach time. While separating the patch set I've most likely stumbled over the problem why this previously raised some lockdep warning with i915, it turned out to be just a might_lock() at the wrong place. Please review and/or comment, Christian.

4 years, 6 months

Re: [Linaro-mm-sig] [PATCH] dma-buf/resv: fix exclusive fence get

by Daniel Vetter

On Sun, Sep 22, 2019 at 2:08 PM Qiang Yu <yuq825(a)gmail.com> wrote: > > This causes kernel crash when testing lima driver. > > Cc: Christian König <christian.koenig(a)amd.com> > Fixes: b8c036dfc66f ("dma-buf: simplify reservation_object_get_fences_rcu a bit") > Signed-off-by: Qiang Yu <yuq825(a)gmail.com> Selftest for this would be lovely, now that the basic infrastructure is in place ... -Daniel > --- > drivers/dma-buf/dma-resv.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/dma-buf/dma-resv.c b/drivers/dma-buf/dma-resv.c > index 42a8f3f11681..709002515550 100644 > --- a/drivers/dma-buf/dma-resv.c > +++ b/drivers/dma-buf/dma-resv.c > @@ -471,7 +471,7 @@ int dma_resv_get_fences_rcu(struct dma_resv *obj, > if (pfence_excl) > *pfence_excl = fence_excl; > else if (fence_excl) > - shared[++shared_count] = fence_excl; > + shared[shared_count++] = fence_excl; > > if (!shared_count) { > kfree(shared); > -- > 2.17.1 > > _______________________________________________ > dri-devel mailing list > dri-devel(a)lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/dri-devel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch

4 years, 6 months

Re: [Linaro-mm-sig] [PATCH] ion_system_heap: support X86 archtecture

by Laura Abbott

On 9/29/19 3:28 AM, jun.zhang(a)intel.com wrote: > From: zhang jun <jun.zhang(a)intel.com> > > we see tons of warning like: > [ 45.846872] x86/PAT: NDK MediaCodec_:3753 map pfn RAM range req > write-combining for [mem 0x1e7a80000-0x1e7a87fff], got write-back > [ 45.848827] x86/PAT: .vorbis.decoder:4088 map pfn RAM range req > write-combining for [mem 0x1e7a58000-0x1e7a58fff], got write-back > [ 45.848875] x86/PAT: NDK MediaCodec_:3753 map pfn RAM range req > write-combining for [mem 0x1e7a48000-0x1e7a4ffff], got write-back > [ 45.849403] x86/PAT: .vorbis.decoder:4088 map pfn RAM range > req write-combining for [mem 0x1e7a70000-0x1e7a70fff], got write-back > > check the kernel Documentation/x86/pat.txt, it says: > A. Exporting pages to users with remap_pfn_range, io_remap_pfn_range, > vm_insert_pfn > Drivers wanting to export some pages to userspace do it by using > mmap interface and a combination of > 1) pgprot_noncached() > 2) io_remap_pfn_range() or remap_pfn_range() or vm_insert_pfn() > With PAT support, a new API pgprot_writecombine is being added. > So, drivers can continue to use the above sequence, with either > pgprot_noncached() or pgprot_writecombine() in step 1, followed by step 2. > > In addition, step 2 internally tracks the region as UC or WC in > memtype list in order to ensure no conflicting mapping. > > Note that this set of APIs only works with IO (non RAM) regions. > If driver ants to export a RAM region, it has to do set_memory_uc() or > set_memory_wc() as step 0 above and also track the usage of those pages > and use set_memory_wb() before the page is freed to free pool. > > the fix follow the pat document, do set_memory_wc() as step 0 and > use the set_memory_wb() before the page is freed. > All this work needs to be done on the new dma-buf heap rework and I don't think it makes sense to put it on the staging version https://lore.kernel.org/lkml/20190906184712.91980-1-john.stultz@linaro.org/ (I also continue to question the value of uncached buffers, especially on x86) > Signed-off-by: he, bo <bo.he(a)intel.com> > Signed-off-by: zhang jun <jun.zhang(a)intel.com> > Signed-off-by: Bai, Jie A <jie.a.bai(a)intel.com> > --- > drivers/staging/android/ion/ion_system_heap.c | 28 ++++++++++++++++++- > 1 file changed, 27 insertions(+), 1 deletion(-) > > diff --git a/drivers/staging/android/ion/ion_system_heap.c b/drivers/staging/android/ion/ion_system_heap.c > index b83a1d16bd89..d298b8194820 100644 > --- a/drivers/staging/android/ion/ion_system_heap.c > +++ b/drivers/staging/android/ion/ion_system_heap.c > @@ -13,6 +13,7 @@ > #include <linux/scatterlist.h> > #include <linux/slab.h> > #include <linux/vmalloc.h> > +#include <asm/set_memory.h> > > #include "ion.h" > > @@ -134,6 +135,13 @@ static int ion_system_heap_allocate(struct ion_heap *heap, > sg = table->sgl; > list_for_each_entry_safe(page, tmp_page, &pages, lru) { > sg_set_page(sg, page, page_size(page), 0); > + > +#ifdef CONFIG_X86 > + if (!(buffer->flags & ION_FLAG_CACHED)) > + set_memory_wc((unsigned long)page_address(sg_page(sg)), > + PAGE_ALIGN(sg->length) >> PAGE_SHIFT); > +#endif > + > sg = sg_next(sg); > list_del(&page->lru); > } > @@ -162,8 +170,15 @@ static void ion_system_heap_free(struct ion_buffer *buffer) > if (!(buffer->private_flags & ION_PRIV_FLAG_SHRINKER_FREE)) > ion_heap_buffer_zero(buffer); > > - for_each_sg(table->sgl, sg, table->nents, i) > + for_each_sg(table->sgl, sg, table->nents, i) { > +#ifdef CONFIG_X86 > + if (!(buffer->flags & ION_FLAG_CACHED)) > + set_memory_wb((unsigned long)page_address(sg_page(sg)), > + PAGE_ALIGN(sg->length) >> PAGE_SHIFT); > +#endif > + > free_buffer_page(sys_heap, buffer, sg_page(sg)); > + } > sg_free_table(table); > kfree(table); > } > @@ -316,6 +331,12 @@ static int ion_system_contig_heap_allocate(struct ion_heap *heap, > > buffer->sg_table = table; > > +#ifdef CONFIG_X86 > + if (!(buffer->flags & ION_FLAG_CACHED)) > + set_memory_wc((unsigned long)page_address(page), > + PAGE_ALIGN(len) >> PAGE_SHIFT); > +#endif > + > return 0; > > free_table: > @@ -334,6 +355,11 @@ static void ion_system_contig_heap_free(struct ion_buffer *buffer) > unsigned long pages = PAGE_ALIGN(buffer->size) >> PAGE_SHIFT; > unsigned long i; > > +#ifdef CONFIG_X86 > + if (!(buffer->flags & ION_FLAG_CACHED)) > + set_memory_wb((unsigned long)page_address(page), pages); > +#endif > + > for (i = 0; i < pages; i++) > __free_page(page + i); > sg_free_table(table); >

4 years, 6 months

Re: [Linaro-mm-sig] [PATCH] ion_system_heap: support X86 archtecture

by Greg KH

On Sun, Sep 29, 2019 at 03:28:41PM +0800, jun.zhang(a)intel.com wrote: > From: zhang jun <jun.zhang(a)intel.com> > > we see tons of warning like: > [ 45.846872] x86/PAT: NDK MediaCodec_:3753 map pfn RAM range req > write-combining for [mem 0x1e7a80000-0x1e7a87fff], got write-back > [ 45.848827] x86/PAT: .vorbis.decoder:4088 map pfn RAM range req > write-combining for [mem 0x1e7a58000-0x1e7a58fff], got write-back > [ 45.848875] x86/PAT: NDK MediaCodec_:3753 map pfn RAM range req > write-combining for [mem 0x1e7a48000-0x1e7a4ffff], got write-back > [ 45.849403] x86/PAT: .vorbis.decoder:4088 map pfn RAM range > req write-combining for [mem 0x1e7a70000-0x1e7a70fff], got write-back > > check the kernel Documentation/x86/pat.txt, it says: > A. Exporting pages to users with remap_pfn_range, io_remap_pfn_range, > vm_insert_pfn > Drivers wanting to export some pages to userspace do it by using > mmap interface and a combination of > 1) pgprot_noncached() > 2) io_remap_pfn_range() or remap_pfn_range() or vm_insert_pfn() > With PAT support, a new API pgprot_writecombine is being added. > So, drivers can continue to use the above sequence, with either > pgprot_noncached() or pgprot_writecombine() in step 1, followed by step 2. > > In addition, step 2 internally tracks the region as UC or WC in > memtype list in order to ensure no conflicting mapping. > > Note that this set of APIs only works with IO (non RAM) regions. > If driver ants to export a RAM region, it has to do set_memory_uc() or > set_memory_wc() as step 0 above and also track the usage of those pages > and use set_memory_wb() before the page is freed to free pool. > > the fix follow the pat document, do set_memory_wc() as step 0 and > use the set_memory_wb() before the page is freed. > > Signed-off-by: he, bo <bo.he(a)intel.com> > Signed-off-by: zhang jun <jun.zhang(a)intel.com> > Signed-off-by: Bai, Jie A <jie.a.bai(a)intel.com> > --- > drivers/staging/android/ion/ion_system_heap.c | 28 ++++++++++++++++++- > 1 file changed, 27 insertions(+), 1 deletion(-) > > diff --git a/drivers/staging/android/ion/ion_system_heap.c b/drivers/staging/android/ion/ion_system_heap.c > index b83a1d16bd89..d298b8194820 100644 > --- a/drivers/staging/android/ion/ion_system_heap.c > +++ b/drivers/staging/android/ion/ion_system_heap.c > @@ -13,6 +13,7 @@ > #include <linux/scatterlist.h> > #include <linux/slab.h> > #include <linux/vmalloc.h> > +#include <asm/set_memory.h> > > #include "ion.h" > > @@ -134,6 +135,13 @@ static int ion_system_heap_allocate(struct ion_heap *heap, > sg = table->sgl; > list_for_each_entry_safe(page, tmp_page, &pages, lru) { > sg_set_page(sg, page, page_size(page), 0); > + > +#ifdef CONFIG_X86 > + if (!(buffer->flags & ION_FLAG_CACHED)) > + set_memory_wc((unsigned long)page_address(sg_page(sg)), > + PAGE_ALIGN(sg->length) >> PAGE_SHIFT); > +#endif There is no way to do this without these #ifdefs? That feels odd, why can't you just always test for this? thanks, greg k-h

4 years, 6 months

Re: [Linaro-mm-sig] [PATCH] dma-buf/resv: fix exclusive fence get

by Chris Wilson

Quoting Qiang Yu (2019-09-22 08:49:00) > This causes kernel crash when testing lima driver. > > Cc: Christian König <christian.koenig(a)amd.com> > Fixes: b8c036dfc66f ("dma-buf: simplify reservation_object_get_fences_rcu a bit") > Signed-off-by: Qiang Yu <yuq825(a)gmail.com> > --- > drivers/dma-buf/dma-resv.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/dma-buf/dma-resv.c b/drivers/dma-buf/dma-resv.c > index 42a8f3f11681..709002515550 100644 > --- a/drivers/dma-buf/dma-resv.c > +++ b/drivers/dma-buf/dma-resv.c > @@ -471,7 +471,7 @@ int dma_resv_get_fences_rcu(struct dma_resv *obj, > if (pfence_excl) > *pfence_excl = fence_excl; > else if (fence_excl) > - shared[++shared_count] = fence_excl; > + shared[shared_count++] = fence_excl; Oops. Reviewed-by: Chris Wilson <chris(a)chris-wilson.co.uk> -Chris

4 years, 6 months

Re: [Linaro-mm-sig] [PATCH] staging: android: ion: Replace strncpy() for stracpy()

by Dan Carpenter

On Sun, Sep 08, 2019 at 02:34:50PM +1000, Adam Zerella wrote: > Using strncpy() does not always terminate the destination string. > stracpy() is a alternative function that does, by using this new > function we will no longer need to insert a null separator. > > Signed-off-by: Adam Zerella <adam.zerella(a)gmail.com> > --- > drivers/staging/android/ion/ion.c | 3 +-- > 1 file changed, 1 insertion(+), 2 deletions(-) > > diff --git a/drivers/staging/android/ion/ion.c b/drivers/staging/android/ion/ion.c > index e6b1ca141b93..17901bd626be 100644 > --- a/drivers/staging/android/ion/ion.c > +++ b/drivers/staging/android/ion/ion.c > @@ -433,8 +433,7 @@ static int ion_query_heaps(struct ion_heap_query *query) > max_cnt = query->cnt; > > plist_for_each_entry(heap, &dev->heaps, node) { > - strncpy(hdata.name, heap->name, MAX_HEAP_NAME); > - hdata.name[sizeof(hdata.name) - 1] = '\0'; > + stracpy(hdata.name, heap->name, MAX_HEAP_NAME); stracpy() only takes two arguments. This doesn't compile. regards, dan carpenter

4 years, 7 months

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

Linaro-mm-sig September 2019