From: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
To: Stefano Stabellini <sstabellini@kernel.org>
Cc: xen-devel@lists.xenproject.org, julien.grall@arm.com,
andrii_anisov@epam.com, oleksandr.dmytryshyn@globallogic.com,
Andrii Anisov <andrii.anisov@gmail.com>
Subject: Re: [PATCH v3 2/2] swiotlb-xen: implement xen_swiotlb_get_sgtable callback
Date: Tue, 31 Jan 2017 14:17:48 -0500 [thread overview]
Message-ID: <20170131191748.GI4949@char.us.ORACLE.com> (raw)
In-Reply-To: <alpine.DEB.2.10.1701311110140.17946@sstabellini-ThinkPad-X260>
On Tue, Jan 31, 2017 at 11:15:50AM -0800, Stefano Stabellini wrote:
> On Tue, 31 Jan 2017, Konrad Rzeszutek Wilk wrote:
> > On Tue, Jan 31, 2017 at 08:30:26PM +0200, Andrii Anisov wrote:
> > > From: Andrii Anisov <andrii_anisov@epam.com>
> > >
> > > Signed-off-by: Andrii Anisov <andrii_anisov@epam.com>
> > > Signed-off-by: Stefano Stabellini <sstabellini@kernel.org>
> > > ---
> > > arch/arm/xen/mm.c | 1 +
> > > drivers/xen/swiotlb-xen.c | 22 ++++++++++++++++++++++
> > > include/xen/swiotlb-xen.h | 6 ++++++
> > > 3 files changed, 29 insertions(+)
> > >
> > > diff --git a/arch/arm/xen/mm.c b/arch/arm/xen/mm.c
> > > index cd1684e..76ea48a 100644
> > > --- a/arch/arm/xen/mm.c
> > > +++ b/arch/arm/xen/mm.c
> > > @@ -199,6 +199,7 @@ static struct dma_map_ops xen_swiotlb_dma_ops = {
> > > .dma_supported = xen_swiotlb_dma_supported,
> > > .set_dma_mask = xen_swiotlb_set_dma_mask,
> > > .mmap = xen_swiotlb_dma_mmap,
> > > + .get_sgtable = xen_swiotlb_get_sgtable,
> > > };
> > >
> > > int __init xen_mm_init(void)
> > > diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
> > > index 8ac36b4..a809d43 100644
> > > --- a/drivers/xen/swiotlb-xen.c
> > > +++ b/drivers/xen/swiotlb-xen.c
> > > @@ -699,3 +699,25 @@ xen_swiotlb_dma_mmap(struct device *dev, struct vm_area_struct *vma,
> > > return dma_common_mmap(dev, vma, cpu_addr, dma_addr, size);
> > > }
> > > EXPORT_SYMBOL_GPL(xen_swiotlb_dma_mmap);
> > > +
> > > +/*
> > > + * Following function should be called with the local pages only.
> >
> > What does 'local pages' mean?
> >
> > > + */
> > > +int
> > > +xen_swiotlb_get_sgtable(struct device *dev, struct sg_table *sgt,
> > > + void *cpu_addr, dma_addr_t handle, size_t size,
> > > + unsigned long attrs)
> > > +{
> > > +#if defined(CONFIG_ARM) || defined(CONFIG_ARM64)
> > > + if (__generic_dma_ops(dev)->get_sgtable) {
> > > +#ifdef DEBUG
> > > + unsigned long bfn = PHYS_PFN(dma_to_phys(dev, handle));
> > > + BUG_ON (!page_is_ram(bfn));
> > > +#endif
> >
> > Could you remove the above please?
>
> I asked him to add it (see
> http://marc.info/?l=xen-devel&m=148461541618751): the reason is that
> this function cannot work on foreign pages. From the commit message
> (d2b7428eb0caa7c66e34b6ac869a43915b294123) it looks like it is supposed
> to be called on buffers returned by dma_alloc_coherent, thus it should
> be safe. However, if that's not the case in any of the drivers, it
> would lead to memory corruptions. The two lines under ifdef DEBUG are an
> nice way to catch this kind of errors. However, given that they cost a
> few cpu cycles more than necessary, I wouldn't enable them in
> production, hence the ifdef DEBUG.
But there are no Kconfig DEBUG - so you may as well just do
#if 0
around the code.
>
>
> > > + return __generic_dma_ops(dev)->get_sgtable(dev, sgt, cpu_addr,
> > > + handle, size, attrs);
> > > + }
> > > +#endif
> > > + return dma_common_get_sgtable(dev, sgt, cpu_addr, handle, size);
> > > +}
> > > +EXPORT_SYMBOL_GPL(xen_swiotlb_get_sgtable);
> > > diff --git a/include/xen/swiotlb-xen.h b/include/xen/swiotlb-xen.h
> > > index 5c8f4c8..c554c23 100644
> > > --- a/include/xen/swiotlb-xen.h
> > > +++ b/include/xen/swiotlb-xen.h
> > > @@ -60,4 +60,10 @@ extern int
> > > xen_swiotlb_dma_mmap(struct device *dev, struct vm_area_struct *vma,
> > > void *cpu_addr, dma_addr_t dma_addr, size_t size,
> > > unsigned long attrs);
> > > +
> > > +extern int
> > > +xen_swiotlb_get_sgtable(struct device *dev, struct sg_table *sgt,
> > > + void *cpu_addr, dma_addr_t handle, size_t size,
> > > + unsigned long attrs);
> >
> > And perhaps fix this to be aligned properly?
> > > +
> > > #endif /* __LINUX_SWIOTLB_XEN_H */
> > > --
> > > 2.7.4
> > >
> >
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
next prev parent reply other threads:[~2017-01-31 19:18 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-01-31 18:30 [PATCH v3 0/2] XEN SWIOTLB dma operations extension Andrii Anisov
2017-01-31 18:30 ` [PATCH v3 1/2] swiotlb-xen: implement xen_swiotlb_dma_mmap callback Andrii Anisov
2017-01-31 18:39 ` Konrad Rzeszutek Wilk
2017-01-31 19:10 ` Stefano Stabellini
2017-02-01 11:46 ` Andrii Anisov
2017-02-01 14:24 ` Konrad Rzeszutek Wilk
2017-01-31 18:30 ` [PATCH v3 2/2] swiotlb-xen: implement xen_swiotlb_get_sgtable callback Andrii Anisov
2017-01-31 18:37 ` Konrad Rzeszutek Wilk
2017-01-31 19:15 ` Stefano Stabellini
2017-01-31 19:17 ` Konrad Rzeszutek Wilk [this message]
2017-01-31 19:21 ` Stefano Stabellini
2017-02-01 10:42 ` Andrii Anisov
2017-02-01 10:57 ` Julien Grall
2017-02-01 10:58 ` Julien Grall
2017-02-01 10:19 ` Andrii Anisov
2017-02-01 10:46 ` Andrii Anisov
2017-02-01 14:25 ` Konrad Rzeszutek Wilk
2017-02-01 14:30 ` Julien Grall
2017-02-01 15:11 ` 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=20170131191748.GI4949@char.us.ORACLE.com \
--to=konrad.wilk@oracle.com \
--cc=andrii.anisov@gmail.com \
--cc=andrii_anisov@epam.com \
--cc=julien.grall@arm.com \
--cc=oleksandr.dmytryshyn@globallogic.com \
--cc=sstabellini@kernel.org \
--cc=xen-devel@lists.xenproject.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
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;
as well as URLs for NNTP newsgroup(s).