From: konrad.wilk@oracle•com (Konrad Rzeszutek Wilk)
To: linux-arm-kernel@lists•infradead.org
Subject: [PATCH v8 13/19] swiotlb-xen: use xen_dma_map/unmap_page, xen_dma_sync_single_for_cpu/device
Date: Wed, 23 Oct 2013 15:51:54 -0400 [thread overview]
Message-ID: <20131023195154.GB28963@phenom.dumpdata.com> (raw)
In-Reply-To: <alpine.DEB.2.02.1310231816180.26077@kaball.uk.xensource.com>
On Wed, Oct 23, 2013 at 06:20:25PM +0100, Stefano Stabellini wrote:
> On Wed, 23 Oct 2013, Konrad Rzeszutek Wilk wrote:
> > On Thu, Oct 17, 2013 at 06:43:28PM +0100, Stefano Stabellini wrote:
> > > Call xen_dma_map_page, xen_dma_unmap_page, xen_dma_sync_single_for_cpu,
> > > xen_dma_sync_single_for_device from swiotlb-xen to ensure cpu/device
> > > coherency of the pages used for DMA, including the ones belonging to the
> > > swiotlb buffer.
> >
> > You lost me.
> >
> > Isn't it the driver's responsibility to do this?
> >
> > Looking at what 'xen_dma_map_page()' does for x86 it looks to add an extra
> > call - page_to_phys - and we ignore it here.
>
> map_page on arm calls the right cache flushes needed to communicate with
> the device. Same with unmap_page.
If this is flushing the cache then I think it makes more sense to do
that without this fancy 'dma_map_page'.
Just call it 'xen_flush_dma_page' and make it a nop on all platforms
except ARM.
> On x86 they are basically nop.
It calls page_to_phys in your patch. That is hardly nop.
> Only map_page and unmap_page know exactly what needs to be done to
> map a page for dma, this is why we need to call them here.
>
>
> > > Signed-off-by: Stefano Stabellini <stefano.stabellini@eu•citrix.com>
> > > ---
> > > drivers/xen/swiotlb-xen.c | 39 +++++++++++++++++++++++++++++++--------
> > > 1 files changed, 31 insertions(+), 8 deletions(-)
> > >
> > > diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
> > > index 189b8db..4221cb5 100644
> > > --- a/drivers/xen/swiotlb-xen.c
> > > +++ b/drivers/xen/swiotlb-xen.c
> > > @@ -378,8 +378,13 @@ dma_addr_t xen_swiotlb_map_page(struct device *dev, struct page *page,
> > > * buffering it.
> > > */
> > > if (dma_capable(dev, dev_addr, size) &&
> > > - !range_straddles_page_boundary(phys, size) && !swiotlb_force)
> > > + !range_straddles_page_boundary(phys, size) && !swiotlb_force) {
> > > + /* we are not interested in the dma_addr returned by
> > > + * xen_dma_map_page, only in the potential cache flushes executed
> > > + * by the function. */
> > > + xen_dma_map_page(dev, page, offset, size, dir, attrs);
> > > return dev_addr;
> > > + }
> > >
> > > /*
> > > * Oh well, have to allocate and map a bounce buffer.
> > > @@ -388,6 +393,8 @@ dma_addr_t xen_swiotlb_map_page(struct device *dev, struct page *page,
> > > if (map == SWIOTLB_MAP_ERROR)
> > > return DMA_ERROR_CODE;
> > >
> > > + xen_dma_map_page(dev, pfn_to_page(map >> PAGE_SHIFT),
> > > + map & ~PAGE_MASK, size, dir, attrs);
> > > dev_addr = xen_phys_to_bus(map);
> > >
> > > /*
> > > @@ -410,12 +417,15 @@ EXPORT_SYMBOL_GPL(xen_swiotlb_map_page);
> > > * whatever the device wrote there.
> > > */
> > > static void xen_unmap_single(struct device *hwdev, dma_addr_t dev_addr,
> > > - size_t size, enum dma_data_direction dir)
> > > + size_t size, enum dma_data_direction dir,
> > > + struct dma_attrs *attrs)
> > > {
> > > phys_addr_t paddr = xen_bus_to_phys(dev_addr);
> > >
> > > BUG_ON(dir == DMA_NONE);
> > >
> > > + xen_dma_unmap_page(hwdev, paddr, size, dir, attrs);
> > > +
> > > /* NOTE: We use dev_addr here, not paddr! */
> > > if (is_xen_swiotlb_buffer(dev_addr)) {
> > > swiotlb_tbl_unmap_single(hwdev, paddr, size, dir);
> > > @@ -438,7 +448,7 @@ void xen_swiotlb_unmap_page(struct device *hwdev, dma_addr_t dev_addr,
> > > size_t size, enum dma_data_direction dir,
> > > struct dma_attrs *attrs)
> > > {
> > > - xen_unmap_single(hwdev, dev_addr, size, dir);
> > > + xen_unmap_single(hwdev, dev_addr, size, dir, attrs);
> > > }
> > > EXPORT_SYMBOL_GPL(xen_swiotlb_unmap_page);
> > >
> > > @@ -461,11 +471,15 @@ xen_swiotlb_sync_single(struct device *hwdev, dma_addr_t dev_addr,
> > >
> > > BUG_ON(dir == DMA_NONE);
> > >
> > > + if (target == SYNC_FOR_CPU)
> > > + xen_dma_sync_single_for_cpu(hwdev, paddr, size, dir);
> > > +
> > > /* NOTE: We use dev_addr here, not paddr! */
> > > - if (is_xen_swiotlb_buffer(dev_addr)) {
> > > + if (is_xen_swiotlb_buffer(dev_addr))
> > > swiotlb_tbl_sync_single(hwdev, paddr, size, dir, target);
> > > - return;
> > > - }
> > > +
> > > + if (target == SYNC_FOR_DEVICE)
> > > + xen_dma_sync_single_for_cpu(hwdev, paddr, size, dir);
> > >
> > > if (dir != DMA_FROM_DEVICE)
> > > return;
> > > @@ -536,8 +550,17 @@ xen_swiotlb_map_sg_attrs(struct device *hwdev, struct scatterlist *sgl,
> > > return DMA_ERROR_CODE;
> > > }
> > > sg->dma_address = xen_phys_to_bus(map);
> > > - } else
> > > + } else {
> > > + /* we are not interested in the dma_addr returned by
> > > + * xen_dma_map_page, only in the potential cache flushes executed
> > > + * by the function. */
> > > + xen_dma_map_page(hwdev, pfn_to_page(paddr >> PAGE_SHIFT),
> > > + paddr & ~PAGE_MASK,
> > > + sg->length,
> > > + dir,
> > > + attrs);
> > > sg->dma_address = dev_addr;
> > > + }
> > > sg_dma_len(sg) = sg->length;
> > > }
> > > return nelems;
> > > @@ -559,7 +582,7 @@ xen_swiotlb_unmap_sg_attrs(struct device *hwdev, struct scatterlist *sgl,
> > > BUG_ON(dir == DMA_NONE);
> > >
> > > for_each_sg(sgl, sg, nelems, i)
> > > - xen_unmap_single(hwdev, sg->dma_address, sg_dma_len(sg), dir);
> > > + xen_unmap_single(hwdev, sg->dma_address, sg_dma_len(sg), dir, attrs);
> > >
> > > }
> > > EXPORT_SYMBOL_GPL(xen_swiotlb_unmap_sg_attrs);
> > > --
> > > 1.7.2.5
> > >
> >
next prev parent reply other threads:[~2013-10-23 19:51 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-10-17 17:42 [PATCH v8 0/19] enable swiotlb-xen on arm and arm64 Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 01/19] arm: make SWIOTLB available Stefano Stabellini
2013-10-18 15:56 ` Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 02/19] arm64: define DMA_ERROR_CODE Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 03/19] arm/xen,arm64/xen: introduce p2m Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 04/19] xen/x86: allow __set_phys_to_machine for autotranslate guests Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 05/19] xen: make xen_create_contiguous_region return the dma address Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 06/19] xen/arm,arm64: enable SWIOTLB_XEN Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 07/19] swiotlb-xen: introduce xen_swiotlb_set_dma_mask Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 08/19] arm/xen: get_dma_ops: return xen_dma_ops if we are running as xen_initial_domain Stefano Stabellini
2013-10-18 12:22 ` Stefano Stabellini
2013-10-18 15:42 ` Russell King - ARM Linux
2013-10-18 15:46 ` Stefano Stabellini
2013-10-18 15:48 ` Russell King - ARM Linux
2013-10-18 15:59 ` Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 09/19] arm64/xen: " Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 10/19] xen: introduce xen_alloc/free_coherent_pages Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 11/19] swiotlb-xen: use xen_alloc/free_coherent_pages Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 12/19] xen: introduce xen_dma_map/unmap_page and xen_dma_sync_single_for_cpu/device Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 13/19] swiotlb-xen: use xen_dma_map/unmap_page, xen_dma_sync_single_for_cpu/device Stefano Stabellini
2013-10-23 14:09 ` Konrad Rzeszutek Wilk
2013-10-23 17:20 ` Stefano Stabellini
2013-10-23 19:51 ` Konrad Rzeszutek Wilk [this message]
2013-10-24 10:44 ` Stefano Stabellini
2013-10-24 13:11 ` Stefano Stabellini
2013-10-25 20:46 ` Konrad Rzeszutek Wilk
2013-10-25 20:44 ` Konrad Rzeszutek Wilk
2013-10-17 17:43 ` [PATCH v8 14/19] ASoC: Samsung: Rename dma_ops by samsung_dma_ops Stefano Stabellini
2013-10-23 14:39 ` Julien Grall
2013-10-17 17:43 ` [PATCH v8 15/19] swiotlb: print a warning when the swiotlb is full Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 16/19] arm, arm64: do not always merge biovec if we are running on Xen Stefano Stabellini
2013-10-18 12:23 ` [PATCH v8 16/19] arm,arm64: " Stefano Stabellini
2013-10-18 14:07 ` Catalin Marinas
2013-10-17 17:43 ` [PATCH v8 17/19] grant-table: call set_phys_to_machine after mapping grant refs Stefano Stabellini
2013-10-23 14:04 ` Konrad Rzeszutek Wilk
2013-10-23 17:15 ` Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 18/19] swiotlb-xen: no need to do execute two bus_to_phys translations in a row Stefano Stabellini
2013-10-23 14:01 ` Konrad Rzeszutek Wilk
2013-10-23 17:34 ` Stefano Stabellini
2013-10-23 19:53 ` Konrad Rzeszutek Wilk
2013-10-24 10:32 ` Stefano Stabellini
2013-10-17 17:43 ` [PATCH v8 19/19] swiotlb-xen: static inline xen_phys_to_bus, xen_bus_to_phys, xen_virt_to_bus and range_straddles_page_boundary Stefano Stabellini
2013-10-23 14:15 ` [PATCH v8 0/19] enable swiotlb-xen on arm and arm64 Konrad Rzeszutek Wilk
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20131023195154.GB28963@phenom.dumpdata.com \
--to=konrad.wilk@oracle$(echo .)com \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox