Re: [PATCH v3] swiotlb: Adjust SWIOTBL bounce buffer size for SEV guests.

From: Konrad Rzeszutek Wilk
Date: Thu Nov 05 2020 - 15:21:03 EST

Next message: pr-tracker-bot: "Re: [GIT PULL] PCI fixes for v5.10"
Previous message: Christian Eggers: "Re: [RFC PATCH net-next 7/9] net: dsa: microchip: ksz9477: add hardware time stamping support"
In reply to: Ashish Kalra: "Re: [PATCH v3] swiotlb: Adjust SWIOTBL bounce buffer size for SEV guests."
Next in thread: Ashish Kalra: "Re: [PATCH v3] swiotlb: Adjust SWIOTBL bounce buffer size for SEV guests."
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Thu, Nov 05, 2020 at 07:38:28PM +0000, Ashish Kalra wrote:
> On Thu, Nov 05, 2020 at 02:06:49PM -0500, Konrad Rzeszutek Wilk wrote:
> > .
> > > > Right, so I am wondering if we can do this better.
> > > >
> > > > That is you are never going to get any 32-bit devices with SEV right? That
> > > > is there is nothing that bounds you to always use the memory below 4GB?
> > > >
> > >
> > > We do support 32-bit PCIe passthrough devices with SEV.
> >
> > Ewww.. Which devices would this be?
>
> That will be difficult to predict as customers could be doing
> passthrough of all kinds of devices.

But SEV is not on some 1990 hardware. It has PCIe, there is no PCI slots in there.

Is it really possible to have a PCIe device that can't do more than 32-bit DMA?

>
> > >
> > > Therefore, we can't just depend on >4G memory for SWIOTLB bounce buffering
> > > when there is I/O pressure, because we do need to support device
> > > passthrough of 32-bit devices.
> >
> > Presumarily there is just a handful of them?
> >
> Again, it will be incorrect to assume this.
>
> > >
> > > Considering this, we believe that this patch needs to adjust/extend
> > > boot-allocation of SWIOTLB and we want to keep it simple to do this
> > > within a range detemined by amount of allocated guest memory.
> >
> > I would prefer to not have to revert this in a year as customers
> > complain about "I paid $$$ and I am wasting half a gig on something
> > I am not using" and giving customers knobs to tweak this instead of
> > doing the right thing from the start.
>
> Currently, we face a lot of situations where we have to tell our
> internal teams/external customers to explicitly increase SWIOTLB buffer
> via the swiotlb parameter on the kernel command line, especially to
> get better I/O performance numbers with SEV.

Presumarily these are 64-bit?

And what devices do you speak off that are actually affected by
this performance? Increasing the SWIOTLB just means we have more
memory, which in mind means you can have _more_ devices in the guest
that won't handle the fact that DMA mapping returns an error.

Not neccessarily that one device suddenly can go faster.

>
> So by having this SWIOTLB size adjustment done implicitly (even using a
> static logic) is a great win-win situation. In other words, having even
> a simple and static default increase of SWIOTLB buffer size for SEV is
> really useful for us.
>
> We can always think of adding all kinds of heuristics to this, but that
> just adds too much complexity without any predictable performance gain.
>
> And to add, the patch extends the SWIOTLB size as an architecture
> specific callback, currently it is a simple and static logic for SEV/x86
> specific, but there is always an option to tweak/extend it with
> additional logic in the future.

Right, and that is what I would like to talk about as I think you
are going to disappear (aka, busy with other stuff) after this patch goes in.

I need to understand this more than "performance" and "internal teams"
requirements to come up with a better way going forward as surely other
platforms will hit the same issue anyhow.

Lets break this down:

How does the performance improve for one single device if you increase the SWIOTLB?
Is there a specific device/driver that you can talk about that improve with this patch?

>
> Thanks,
> Ashish'
>
> >
> > That is the right thing being something less static.
> >
> > Can you work with me on what that could be please?
> >
> > >
> > > Thanks,
> > > Ashish
> > >
> > > > What I wonder is if we can combine the boot-allocation of the SWIOTLB
> > > > with the post-boot-allocation of SWIOLTB to stitch together
> > > > continous physical ranges.
> > > >
> > > > That way you have the flexibility at the start of using 64MB but if there
> > > > is pressure, we grow to a bigger size?
> > > >
> > > > >
> > > > > Thanks,
> > > > > Ashish
> > > > >
> > > > > > > memory.
> > > > > > >
> > > > > > > Using late_initcall() interface to invoke
> > > > > > > swiotlb_adjust() does not work as the size
> > > > > > > adjustment needs to be done before mem_encrypt_init()
> > > > > > > and reserve_crashkernel() which use the allocated
> > > > > > > SWIOTLB buffer size, hence calling it explicitly
> > > > > > > from setup_arch().
> > > > > > >
> > > > > > > The SWIOTLB default size adjustment is added as an
> > > > > > > architecture specific interface/callback to allow
> > > > > > > architectures such as those supporting memory
> > > > > > > encryption to adjust/expand SWIOTLB size for their
> > > > > > > use.
> > > > > > >
> > > > > > > Signed-off-by: Ashish Kalra <ashish.kalra@xxxxxxx>
> > > > > > > ---
> > > > > > > arch/x86/kernel/setup.c | 2 ++
> > > > > > > arch/x86/mm/mem_encrypt.c | 42 +++++++++++++++++++++++++++++++++++++++
> > > > > > > include/linux/swiotlb.h | 1 +
> > > > > > > kernel/dma/swiotlb.c | 27 +++++++++++++++++++++++++
> > > > > > > 4 files changed, 72 insertions(+)
> > > > > > >
> > > > > > > diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c
> > > > > > > index 3511736fbc74..b073d58dd4a3 100644
> > > > > > > --- a/arch/x86/kernel/setup.c
> > > > > > > +++ b/arch/x86/kernel/setup.c
> > > > > > > @@ -1166,6 +1166,8 @@ void __init setup_arch(char **cmdline_p)
> > > > > > > if (boot_cpu_has(X86_FEATURE_GBPAGES))
> > > > > > > hugetlb_cma_reserve(PUD_SHIFT - PAGE_SHIFT);
> > > > > > >
> > > > > > > + swiotlb_adjust();
> > > > > > > +
> > > > > > > /*
> > > > > > > * Reserve memory for crash kernel after SRAT is parsed so that it
> > > > > > > * won't consume hotpluggable memory.
> > > > > > > diff --git a/arch/x86/mm/mem_encrypt.c b/arch/x86/mm/mem_encrypt.c
> > > > > > > index 3f248f0d0e07..e0deb157cddd 100644
> > > > > > > --- a/arch/x86/mm/mem_encrypt.c
> > > > > > > +++ b/arch/x86/mm/mem_encrypt.c
> > > > > > > @@ -489,7 +489,49 @@ static void print_mem_encrypt_feature_info(void)
> > > > > > > pr_cont("\n");
> > > > > > > }
> > > > > > >
> > > > > > > +#define TOTAL_MEM_1G 0x40000000UL
> > > > > > > +#define TOTAL_MEM_4G 0x100000000UL
> > > > > > > +
> > > > > > > +#define SIZE_128M (128UL<<20)
> > > > > > > +#define SIZE_256M (256UL<<20)
> > > > > > > +#define SIZE_512M (512UL<<20)
> > > > > > > +
> > > > > > > /* Architecture __weak replacement functions */
> > > > > > > +unsigned long __init arch_swiotlb_adjust(unsigned long iotlb_default_size)
> > > > > > > +{
> > > > > > > + unsigned long size = 0;
> > > > > > > +
> > > > > > > + /*
> > > > > > > + * For SEV, all DMA has to occur via shared/unencrypted pages.
> > > > > > > + * SEV uses SWOTLB to make this happen without changing device
> > > > > > > + * drivers. However, depending on the workload being run, the
> > > > > > > + * default 64MB of SWIOTLB may not be enough & SWIOTLB may
> > > > > > > + * run out of buffers for DMA, resulting in I/O errors and/or
> > > > > > > + * performance degradation especially with high I/O workloads.
> > > > > > > + * Increase the default size of SWIOTLB for SEV guests using
> > > > > > > + * a minimum value of 128MB and a maximum value of 512MB,
> > > > > > > + * depending on amount of provisioned guest memory.
> > > > > > > + */
> > > > > > > + if (sev_active()) {
> > > > > > > + phys_addr_t total_mem = memblock_phys_mem_size();
> > > > > > > +
> > > > > > > + if (total_mem <= TOTAL_MEM_1G)
> > > > > > > + size = clamp(iotlb_default_size * 2, SIZE_128M,
> > > > > > > + SIZE_128M);
> > > > > > > + else if (total_mem <= TOTAL_MEM_4G)
> > > > > > > + size = clamp(iotlb_default_size * 4, SIZE_256M,
> > > > > > > + SIZE_256M);
> > > > > > > + else
> > > > > > > + size = clamp(iotlb_default_size * 8, SIZE_512M,
> > > > > > > + SIZE_512M);
> > > > > > > +
> > > > > > > + pr_info("SEV adjusted max SWIOTLB size = %luMB",
> > > > > > > + size >> 20);
> > > > > > > + }
> > > > > > > +
> > > > > > > + return size;
> > > > > > > +}
> > > > > > > +
> > > > > > > void __init mem_encrypt_init(void)
> > > > > > > {
> > > > > > > if (!sme_me_mask)
> > > > > > > diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
> > > > > > > index 046bb94bd4d6..01ae6d891327 100644
> > > > > > > --- a/include/linux/swiotlb.h
> > > > > > > +++ b/include/linux/swiotlb.h
> > > > > > > @@ -33,6 +33,7 @@ extern void swiotlb_init(int verbose);
> > > > > > > int swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose);
> > > > > > > extern unsigned long swiotlb_nr_tbl(void);
> > > > > > > unsigned long swiotlb_size_or_default(void);
> > > > > > > +extern void __init swiotlb_adjust(void);
> > > > > > > extern int swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs);
> > > > > > > extern void __init swiotlb_update_mem_attributes(void);
> > > > > > >
> > > > > > > diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
> > > > > > > index c19379fabd20..66a9e627bb51 100644
> > > > > > > --- a/kernel/dma/swiotlb.c
> > > > > > > +++ b/kernel/dma/swiotlb.c
> > > > > > > @@ -163,6 +163,33 @@ unsigned long swiotlb_size_or_default(void)
> > > > > > > return size ? size : (IO_TLB_DEFAULT_SIZE);
> > > > > > > }
> > > > > > >
> > > > > > > +unsigned long __init __weak arch_swiotlb_adjust(unsigned long size)
> > > > > > > +{
> > > > > > > + return 0;
> > > > > > > +}
> > > > > > > +
> > > > > > > +void __init swiotlb_adjust(void)
> > > > > > > +{
> > > > > > > + unsigned long size;
> > > > > > > +
> > > > > > > + /*
> > > > > > > + * If swiotlb parameter has not been specified, give a chance to
> > > > > > > + * architectures such as those supporting memory encryption to
> > > > > > > + * adjust/expand SWIOTLB size for their use.
> > > > > > > + */
> > > > > > > + if (!io_tlb_nslabs) {
> > > > > > > + size = arch_swiotlb_adjust(IO_TLB_DEFAULT_SIZE);
> > > > > > > + if (size) {
> > > > > > > + size = ALIGN(size, 1 << IO_TLB_SHIFT);
> > > > > > > + io_tlb_nslabs = size >> IO_TLB_SHIFT;
> > > > > > > + io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
> > > > > > > +
> > > > > > > + pr_info("architecture adjusted SWIOTLB slabs = %lu\n",
> > > > > > > + io_tlb_nslabs);
> > > > > > > + }
> > > > > > > + }
> > > > > > > +}
> > > > > > > +
> > > > > > > void swiotlb_print_info(void)
> > > > > > > {
> > > > > > > unsigned long bytes = io_tlb_nslabs << IO_TLB_SHIFT;
> > > > > > > --
> > > > > > > 2.17.1
> > > > > > >

Next message: pr-tracker-bot: "Re: [GIT PULL] PCI fixes for v5.10"
Previous message: Christian Eggers: "Re: [RFC PATCH net-next 7/9] net: dsa: microchip: ksz9477: add hardware time stamping support"
In reply to: Ashish Kalra: "Re: [PATCH v3] swiotlb: Adjust SWIOTBL bounce buffer size for SEV guests."
Next in thread: Ashish Kalra: "Re: [PATCH v3] swiotlb: Adjust SWIOTBL bounce buffer size for SEV guests."
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]