* [PATCH V2] mm/memtest: Add ARCH_USE_MEMTEST
From: Anshuman Khandual @ 2021-03-01 4:32 UTC (permalink / raw)
To: linux-mm
Cc: Chris Zankel, Thomas Bogendoerfer, Anshuman Khandual,
linux-xtensa, linuxppc-dev, linux-kernel, Russell King,
linux-mips, Max Filippov, Ingo Molnar, Paul Mackerras,
Catalin Marinas, Thomas Gleixner, Will Deacon, linux-arm-kernel
early_memtest() does not get called from all architectures. Hence enabling
CONFIG_MEMTEST and providing a valid memtest=[1..N] kernel command line
option might not trigger the memory pattern tests as would be expected in
normal circumstances. This situation is misleading.
The change here prevents the above mentioned problem after introducing a
new config option ARCH_USE_MEMTEST that should be subscribed on platforms
that call early_memtest(), in order to enable the config CONFIG_MEMTEST.
Conversely CONFIG_MEMTEST cannot be enabled on platforms where it would
not be tested anyway.
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Chris Zankel <chris@zankel.net>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-mips@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: linux-xtensa@linux-xtensa.org
Cc: linux-mm@kvack.org
Cc: linux-kernel@vger.kernel.org
Reviewed-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
This patch applies on v5.12-rc1 and has been tested on arm64 platform.
But it has been just build tested on all other platforms.
Changes in V2:
- Added ARCH_USE_MEMTEST in the sorted alphabetical order on platforms
Changes in V1:
https://patchwork.kernel.org/project/linux-mm/patch/1612498242-31579-1-git-send-email-anshuman.khandual@arm.com/
arch/arm/Kconfig | 1 +
arch/arm64/Kconfig | 1 +
arch/mips/Kconfig | 1 +
arch/powerpc/Kconfig | 1 +
arch/x86/Kconfig | 1 +
arch/xtensa/Kconfig | 1 +
lib/Kconfig.debug | 9 ++++++++-
7 files changed, 14 insertions(+), 1 deletion(-)
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index 853aab5ab327..9ab047d4cd0a 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -33,6 +33,7 @@ config ARM
select ARCH_SUPPORTS_ATOMIC_RMW
select ARCH_USE_BUILTIN_BSWAP
select ARCH_USE_CMPXCHG_LOCKREF
+ select ARCH_USE_MEMTEST
select ARCH_WANT_DEFAULT_TOPDOWN_MMAP_LAYOUT if MMU
select ARCH_WANT_IPC_PARSE_VERSION
select ARCH_WANT_LD_ORPHAN_WARN
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig
index 1f212b47a48a..d4fe5118e9c8 100644
--- a/arch/arm64/Kconfig
+++ b/arch/arm64/Kconfig
@@ -67,6 +67,7 @@ config ARM64
select ARCH_KEEP_MEMBLOCK
select ARCH_USE_CMPXCHG_LOCKREF
select ARCH_USE_GNU_PROPERTY
+ select ARCH_USE_MEMTEST
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_USE_QUEUED_SPINLOCKS
select ARCH_USE_SYM_ANNOTATIONS
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index d89efba3d8a4..93a4f502f962 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -14,6 +14,7 @@ config MIPS
select ARCH_SUPPORTS_UPROBES
select ARCH_USE_BUILTIN_BSWAP
select ARCH_USE_CMPXCHG_LOCKREF if 64BIT
+ select ARCH_USE_MEMTEST
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_USE_QUEUED_SPINLOCKS
select ARCH_WANT_DEFAULT_TOPDOWN_MMAP_LAYOUT if MMU
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 386ae12d8523..3778ad17f56a 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -149,6 +149,7 @@ config PPC
select ARCH_SUPPORTS_DEBUG_PAGEALLOC if PPC32 || PPC_BOOK3S_64
select ARCH_USE_BUILTIN_BSWAP
select ARCH_USE_CMPXCHG_LOCKREF if PPC64
+ select ARCH_USE_MEMTEST
select ARCH_USE_QUEUED_RWLOCKS if PPC_QUEUED_SPINLOCKS
select ARCH_USE_QUEUED_SPINLOCKS if PPC_QUEUED_SPINLOCKS
select ARCH_WANT_IPC_PARSE_VERSION
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 2792879d398e..2cb76fd5258e 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -100,6 +100,7 @@ config X86
select ARCH_SUPPORTS_LTO_CLANG if X86_64
select ARCH_SUPPORTS_LTO_CLANG_THIN if X86_64
select ARCH_USE_BUILTIN_BSWAP
+ select ARCH_USE_MEMTEST
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_USE_QUEUED_SPINLOCKS
select ARCH_USE_SYM_ANNOTATIONS
diff --git a/arch/xtensa/Kconfig b/arch/xtensa/Kconfig
index a99dc39f6964..ca51896c53df 100644
--- a/arch/xtensa/Kconfig
+++ b/arch/xtensa/Kconfig
@@ -7,6 +7,7 @@ config XTENSA
select ARCH_HAS_SYNC_DMA_FOR_CPU if MMU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE if MMU
select ARCH_HAS_DMA_SET_UNCACHED if MMU
+ select ARCH_USE_MEMTEST
select ARCH_USE_QUEUED_RWLOCKS
select ARCH_USE_QUEUED_SPINLOCKS
select ARCH_WANT_FRAME_POINTERS
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index a2d04c00cda2..2c296535a4b3 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -2521,11 +2521,18 @@ config TEST_FPU
endif # RUNTIME_TESTING_MENU
+config ARCH_USE_MEMTEST
+ bool
+ help
+ An architecture should select this when it uses early_memtest()
+ during boot process.
+
config MEMTEST
bool "Memtest"
+ depends on ARCH_USE_MEMTEST
help
This option adds a kernel parameter 'memtest', which allows memtest
- to be set.
+ to be set and executed.
memtest=0, mean disabled; -- default
memtest=1, mean do 1 test pattern;
...
--
2.20.1
^ permalink raw reply related
* [PATCH] mm: Generalize HUGETLB_PAGE_SIZE_VARIABLE
From: Anshuman Khandual @ 2021-03-01 5:50 UTC (permalink / raw)
To: linux-mm
Cc: linux-ia64, Anshuman Khandual, linux-kernel, Paul Mackerras,
Andrew Morton, linuxppc-dev, Christoph Hellwig
HUGETLB_PAGE_SIZE_VARIABLE need not be defined for each individual
platform subscribing it. Instead just make it generic.
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: linux-ia64@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: linux-mm@kvack.org
Cc: linux-kernel@vger.kernel.org
Suggested-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
This change was originally suggested in an earilier discussion. This
applies on v5.12-rc1 and has been build tested on all applicable
platforms i.e ia64 and powerpc.
https://patchwork.kernel.org/project/linux-mm/patch/1613024531-19040-3-git-send-email-anshuman.khandual@arm.com/
arch/ia64/Kconfig | 6 +-----
arch/powerpc/Kconfig | 6 +-----
mm/Kconfig | 8 ++++++++
3 files changed, 10 insertions(+), 10 deletions(-)
diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
index 2ad7a8d29fcc..6b3e3f6c29ae 100644
--- a/arch/ia64/Kconfig
+++ b/arch/ia64/Kconfig
@@ -32,6 +32,7 @@ config IA64
select TTY
select HAVE_ARCH_TRACEHOOK
select HAVE_VIRT_CPU_ACCOUNTING
+ select HUGETLB_PAGE_SIZE_VARIABLE
select VIRT_TO_BUS
select GENERIC_IRQ_PROBE
select GENERIC_PENDING_IRQ if SMP
@@ -82,11 +83,6 @@ config STACKTRACE_SUPPORT
config GENERIC_LOCKBREAK
def_bool n
-config HUGETLB_PAGE_SIZE_VARIABLE
- bool
- depends on HUGETLB_PAGE
- default y
-
config GENERIC_CALIBRATE_DELAY
bool
default y
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 3778ad17f56a..b8565bed284f 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -232,6 +232,7 @@ config PPC
select HAVE_HARDLOCKUP_DETECTOR_PERF if PERF_EVENTS && HAVE_PERF_EVENTS_NMI && !HAVE_HARDLOCKUP_DETECTOR_ARCH
select HAVE_PERF_REGS
select HAVE_PERF_USER_STACK_DUMP
+ select HUGETLB_PAGE_SIZE_VARIABLE if PPC_BOOK3S_64
select MMU_GATHER_RCU_TABLE_FREE
select MMU_GATHER_PAGE_SIZE
select HAVE_REGS_AND_STACK_ACCESS_API
@@ -416,11 +417,6 @@ config HIGHMEM
source "kernel/Kconfig.hz"
-config HUGETLB_PAGE_SIZE_VARIABLE
- bool
- depends on HUGETLB_PAGE && PPC_BOOK3S_64
- default y
-
config MATH_EMULATION
bool "Math emulation"
depends on 4xx || PPC_8xx || PPC_MPC832x || BOOKE
diff --git a/mm/Kconfig b/mm/Kconfig
index 24c045b24b95..e604a87862a4 100644
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -274,6 +274,14 @@ config ARCH_ENABLE_HUGEPAGE_MIGRATION
config ARCH_ENABLE_THP_MIGRATION
bool
+config HUGETLB_PAGE_SIZE_VARIABLE
+ def_bool n
+ depends on HUGETLB_PAGE
+ help
+ When there are multiple HugeTLB sizes available on a platform
+ and pageblock_order could then be a dynamic value instead of
+ standard HUGETLB_PAGE_ORDER.
+
config CONTIG_ALLOC
def_bool (MEMORY_ISOLATION && COMPACTION) || CMA
--
2.20.1
^ permalink raw reply related
* Re: [PATCH] mm: Generalize HUGETLB_PAGE_SIZE_VARIABLE
From: Christoph Hellwig @ 2021-03-01 6:23 UTC (permalink / raw)
To: Anshuman Khandual
Cc: linux-ia64, linux-kernel, linux-mm, Paul Mackerras, Andrew Morton,
linuxppc-dev, Christoph Hellwig
In-Reply-To: <1614577853-7452-1-git-send-email-anshuman.khandual@arm.com>
On Mon, Mar 01, 2021 at 11:20:53AM +0530, Anshuman Khandual wrote:
> HUGETLB_PAGE_SIZE_VARIABLE need not be defined for each individual
> platform subscribing it. Instead just make it generic.
>
> Cc: Michael Ellerman <mpe@ellerman.id.au>
> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> Cc: Paul Mackerras <paulus@samba.org>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: Christoph Hellwig <hch@lst.de>
> Cc: linux-ia64@vger.kernel.org
> Cc: linuxppc-dev@lists.ozlabs.org
> Cc: linux-mm@kvack.org
> Cc: linux-kernel@vger.kernel.org
> Suggested-by: Christoph Hellwig <hch@lst.de>
> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
> ---
> This change was originally suggested in an earilier discussion. This
> applies on v5.12-rc1 and has been build tested on all applicable
> platforms i.e ia64 and powerpc.
>
> https://patchwork.kernel.org/project/linux-mm/patch/1613024531-19040-3-git-send-email-anshuman.khandual@arm.com/
>
> arch/ia64/Kconfig | 6 +-----
> arch/powerpc/Kconfig | 6 +-----
> mm/Kconfig | 8 ++++++++
> 3 files changed, 10 insertions(+), 10 deletions(-)
>
> diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
> index 2ad7a8d29fcc..6b3e3f6c29ae 100644
> --- a/arch/ia64/Kconfig
> +++ b/arch/ia64/Kconfig
> @@ -32,6 +32,7 @@ config IA64
> select TTY
> select HAVE_ARCH_TRACEHOOK
> select HAVE_VIRT_CPU_ACCOUNTING
> + select HUGETLB_PAGE_SIZE_VARIABLE
doesn't this need a 'if HUGETLB_PAGE'
or did you verify that HUGETLB_PAGE_SIZE_VARIABLE checks are always
nested inside of HUGETLB_PAGE ones?
^ permalink raw reply
* [PATCH kernel v2] powerpc/iommu: Annotate nested lock for lockdep
From: Alexey Kardashevskiy @ 2021-03-01 6:36 UTC (permalink / raw)
To: linuxppc-dev; +Cc: Alexey Kardashevskiy, Frederic Barrat, kvm-ppc
The IOMMU table is divided into pools for concurrent mappings and each
pool has a separate spinlock. When taking the ownership of an IOMMU group
to pass through a device to a VM, we lock these spinlocks which triggers
a false negative warning in lockdep (below).
This fixes it by annotating the large pool's spinlock as a nest lock
which makes lockdep not complaining when locking nested locks if
the nest lock is locked already.
===
WARNING: possible recursive locking detected
5.11.0-le_syzkaller_a+fstn1 #100 Not tainted
--------------------------------------------
qemu-system-ppc/4129 is trying to acquire lock:
c0000000119bddb0 (&(p->lock)/1){....}-{2:2}, at: iommu_take_ownership+0xac/0x1e0
but task is already holding lock:
c0000000119bdd30 (&(p->lock)/1){....}-{2:2}, at: iommu_take_ownership+0xac/0x1e0
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(&(p->lock)/1);
lock(&(p->lock)/1);
===
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
---
Changes:
v2:
* fixed iommu_release_ownership() as well
---
arch/powerpc/kernel/iommu.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/powerpc/kernel/iommu.c b/arch/powerpc/kernel/iommu.c
index c1a5c366a664..d0df3e5ff5e0 100644
--- a/arch/powerpc/kernel/iommu.c
+++ b/arch/powerpc/kernel/iommu.c
@@ -1089,7 +1089,7 @@ int iommu_take_ownership(struct iommu_table *tbl)
spin_lock_irqsave(&tbl->large_pool.lock, flags);
for (i = 0; i < tbl->nr_pools; i++)
- spin_lock(&tbl->pools[i].lock);
+ spin_lock_nest_lock(&tbl->pools[i].lock, &tbl->large_pool.lock);
iommu_table_release_pages(tbl);
@@ -1117,7 +1117,7 @@ void iommu_release_ownership(struct iommu_table *tbl)
spin_lock_irqsave(&tbl->large_pool.lock, flags);
for (i = 0; i < tbl->nr_pools; i++)
- spin_lock(&tbl->pools[i].lock);
+ spin_lock_nest_lock(&tbl->pools[i].lock, &tbl->large_pool.lock);
memset(tbl->it_map, 0, sz);
--
2.17.1
^ permalink raw reply related
* Re: [PATCH] mm: Generalize HUGETLB_PAGE_SIZE_VARIABLE
From: Anshuman Khandual @ 2021-03-01 7:43 UTC (permalink / raw)
To: Christoph Hellwig
Cc: linux-ia64, linux-kernel, linux-mm, Paul Mackerras, Andrew Morton,
linuxppc-dev
In-Reply-To: <20210301062358.GA25761@lst.de>
On 3/1/21 11:53 AM, Christoph Hellwig wrote:
> On Mon, Mar 01, 2021 at 11:20:53AM +0530, Anshuman Khandual wrote:
>> HUGETLB_PAGE_SIZE_VARIABLE need not be defined for each individual
>> platform subscribing it. Instead just make it generic.
>>
>> Cc: Michael Ellerman <mpe@ellerman.id.au>
>> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
>> Cc: Paul Mackerras <paulus@samba.org>
>> Cc: Andrew Morton <akpm@linux-foundation.org>
>> Cc: Christoph Hellwig <hch@lst.de>
>> Cc: linux-ia64@vger.kernel.org
>> Cc: linuxppc-dev@lists.ozlabs.org
>> Cc: linux-mm@kvack.org
>> Cc: linux-kernel@vger.kernel.org
>> Suggested-by: Christoph Hellwig <hch@lst.de>
>> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
>> ---
>> This change was originally suggested in an earilier discussion. This
>> applies on v5.12-rc1 and has been build tested on all applicable
>> platforms i.e ia64 and powerpc.
>>
>> https://patchwork.kernel.org/project/linux-mm/patch/1613024531-19040-3-git-send-email-anshuman.khandual@arm.com/
>>
>> arch/ia64/Kconfig | 6 +-----
>> arch/powerpc/Kconfig | 6 +-----
>> mm/Kconfig | 8 ++++++++
>> 3 files changed, 10 insertions(+), 10 deletions(-)
>>
>> diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig
>> index 2ad7a8d29fcc..6b3e3f6c29ae 100644
>> --- a/arch/ia64/Kconfig
>> +++ b/arch/ia64/Kconfig
>> @@ -32,6 +32,7 @@ config IA64
>> select TTY
>> select HAVE_ARCH_TRACEHOOK
>> select HAVE_VIRT_CPU_ACCOUNTING
>> + select HUGETLB_PAGE_SIZE_VARIABLE
>
> doesn't this need a 'if HUGETLB_PAGE'
While making HUGETLB_PAGE_SIZE_VARIABLE a generic option, also made it
dependent on HUGETLB_PAGE. Should not that gate HUGETLB_PAGE_SIZE_VARIABLE
when HUGETLB_PAGE is not available irrespective of the select statement on
the platforms ?
>
> or did you verify that HUGETLB_PAGE_SIZE_VARIABLE checks are always
> nested inside of HUGETLB_PAGE ones?
>
^ permalink raw reply
* [PATCH 01/14] powerpc/svm: stop using io_tlb_start
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Use the local variable that is passed to swiotlb_init_with_tbl for
freeing the memory in the failure case to isolate the code a little
better from swiotlb internals.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
arch/powerpc/platforms/pseries/svm.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/powerpc/platforms/pseries/svm.c b/arch/powerpc/platforms/pseries/svm.c
index 7b739cc7a8a93e..1d829e257996fb 100644
--- a/arch/powerpc/platforms/pseries/svm.c
+++ b/arch/powerpc/platforms/pseries/svm.c
@@ -55,9 +55,9 @@ void __init svm_swiotlb_init(void)
if (vstart && !swiotlb_init_with_tbl(vstart, io_tlb_nslabs, false))
return;
- if (io_tlb_start)
- memblock_free_early(io_tlb_start,
- PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT));
+
+ memblock_free_early(__pa(vstart),
+ PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT));
panic("SVM: Cannot allocate SWIOTLB buffer");
}
--
2.29.2
^ permalink raw reply related
* swiotlb cleanups v2
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
Hi Konrad,
this series contains a bunch of swiotlb cleanups, mostly to reduce the
amount of internals exposed to code outside of swiotlb.c, which should
helper to prepare for supporting multiple different bounce buffer pools.
Changes since v1:
- rebased to v5.12-rc1
- a few more cleanups
- merge and forward port the patch from Claire to move all the global
variables into a struct to prepare for multiple instances
^ permalink raw reply
* [PATCH 02/14] swiotlb: remove the alloc_size parameter to swiotlb_tbl_unmap_single
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Now that swiotlb remembers the allocation size there is no need to pass
it back to swiotlb_tbl_unmap_single.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/iommu/dma-iommu.c | 11 +++-------
drivers/xen/swiotlb-xen.c | 4 ++--
include/linux/swiotlb.h | 1 -
kernel/dma/direct.h | 2 +-
kernel/dma/swiotlb.c | 45 ++++++++++++++++++++-------------------
5 files changed, 29 insertions(+), 34 deletions(-)
diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
index 9ab6ee22c11088..da2bd8f0885e6e 100644
--- a/drivers/iommu/dma-iommu.c
+++ b/drivers/iommu/dma-iommu.c
@@ -493,8 +493,6 @@ static void __iommu_dma_unmap_swiotlb(struct device *dev, dma_addr_t dma_addr,
unsigned long attrs)
{
struct iommu_domain *domain = iommu_get_dma_domain(dev);
- struct iommu_dma_cookie *cookie = domain->iova_cookie;
- struct iova_domain *iovad = &cookie->iovad;
phys_addr_t phys;
phys = iommu_iova_to_phys(domain, dma_addr);
@@ -504,8 +502,7 @@ static void __iommu_dma_unmap_swiotlb(struct device *dev, dma_addr_t dma_addr,
__iommu_dma_unmap(dev, dma_addr, size);
if (unlikely(is_swiotlb_buffer(phys)))
- swiotlb_tbl_unmap_single(dev, phys, size,
- iova_align(iovad, size), dir, attrs);
+ swiotlb_tbl_unmap_single(dev, phys, size, dir, attrs);
}
static bool dev_is_untrusted(struct device *dev)
@@ -580,10 +577,8 @@ static dma_addr_t __iommu_dma_map_swiotlb(struct device *dev, phys_addr_t phys,
}
iova = __iommu_dma_map(dev, phys, aligned_size, prot, dma_mask);
- if ((iova == DMA_MAPPING_ERROR) && is_swiotlb_buffer(phys))
- swiotlb_tbl_unmap_single(dev, phys, org_size,
- aligned_size, dir, attrs);
-
+ if (iova == DMA_MAPPING_ERROR && is_swiotlb_buffer(phys))
+ swiotlb_tbl_unmap_single(dev, phys, org_size, dir, attrs);
return iova;
}
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 2b385c1b4a99cb..d47f1b311caac0 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -406,7 +406,7 @@ static dma_addr_t xen_swiotlb_map_page(struct device *dev, struct page *page,
* Ensure that the address returned is DMA'ble
*/
if (unlikely(!dma_capable(dev, dev_addr, size, true))) {
- swiotlb_tbl_unmap_single(dev, map, size, size, dir,
+ swiotlb_tbl_unmap_single(dev, map, size, dir,
attrs | DMA_ATTR_SKIP_CPU_SYNC);
return DMA_MAPPING_ERROR;
}
@@ -445,7 +445,7 @@ static void xen_swiotlb_unmap_page(struct device *hwdev, dma_addr_t dev_addr,
/* NOTE: We use dev_addr here, not paddr! */
if (is_xen_swiotlb_buffer(hwdev, dev_addr))
- swiotlb_tbl_unmap_single(hwdev, paddr, size, size, dir, attrs);
+ swiotlb_tbl_unmap_single(hwdev, paddr, size, dir, attrs);
}
static void
diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
index 5857a937c63722..59f421d041ed9e 100644
--- a/include/linux/swiotlb.h
+++ b/include/linux/swiotlb.h
@@ -57,7 +57,6 @@ phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t phys,
extern void swiotlb_tbl_unmap_single(struct device *hwdev,
phys_addr_t tlb_addr,
size_t mapping_size,
- size_t alloc_size,
enum dma_data_direction dir,
unsigned long attrs);
diff --git a/kernel/dma/direct.h b/kernel/dma/direct.h
index b9861557873768..e1bf721591c0cf 100644
--- a/kernel/dma/direct.h
+++ b/kernel/dma/direct.h
@@ -114,6 +114,6 @@ static inline void dma_direct_unmap_page(struct device *dev, dma_addr_t addr,
dma_direct_sync_single_for_cpu(dev, addr, size, dir);
if (unlikely(is_swiotlb_buffer(phys)))
- swiotlb_tbl_unmap_single(dev, phys, size, size, dir, attrs);
+ swiotlb_tbl_unmap_single(dev, phys, size, dir, attrs);
}
#endif /* _KERNEL_DMA_DIRECT_H */
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index c10e855a03bc16..03aa614565e417 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -102,7 +102,7 @@ static phys_addr_t *io_tlb_orig_addr;
/*
* The mapped buffer's size should be validated during a sync operation.
*/
-static size_t *io_tlb_orig_size;
+static size_t *io_tlb_alloc_size;
/*
* Protect the above data structures in the map and unmap calls
@@ -253,15 +253,15 @@ int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose)
__func__, alloc_size, PAGE_SIZE);
alloc_size = PAGE_ALIGN(io_tlb_nslabs * sizeof(size_t));
- io_tlb_orig_size = memblock_alloc(alloc_size, PAGE_SIZE);
- if (!io_tlb_orig_size)
+ io_tlb_alloc_size = memblock_alloc(alloc_size, PAGE_SIZE);
+ if (!io_tlb_alloc_size)
panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
__func__, alloc_size, PAGE_SIZE);
for (i = 0; i < io_tlb_nslabs; i++) {
io_tlb_list[i] = IO_TLB_SEGSIZE - io_tlb_offset(i);
io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
- io_tlb_orig_size[i] = 0;
+ io_tlb_alloc_size[i] = 0;
}
io_tlb_index = 0;
no_iotlb_memory = false;
@@ -393,18 +393,18 @@ swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
if (!io_tlb_orig_addr)
goto cleanup4;
- io_tlb_orig_size = (size_t *)
+ io_tlb_alloc_size = (size_t *)
__get_free_pages(GFP_KERNEL,
get_order(io_tlb_nslabs *
sizeof(size_t)));
- if (!io_tlb_orig_size)
+ if (!io_tlb_alloc_size)
goto cleanup5;
for (i = 0; i < io_tlb_nslabs; i++) {
io_tlb_list[i] = IO_TLB_SEGSIZE - io_tlb_offset(i);
io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
- io_tlb_orig_size[i] = 0;
+ io_tlb_alloc_size[i] = 0;
}
io_tlb_index = 0;
no_iotlb_memory = false;
@@ -436,7 +436,7 @@ void __init swiotlb_exit(void)
return;
if (late_alloc) {
- free_pages((unsigned long)io_tlb_orig_size,
+ free_pages((unsigned long)io_tlb_alloc_size,
get_order(io_tlb_nslabs * sizeof(size_t)));
free_pages((unsigned long)io_tlb_orig_addr,
get_order(io_tlb_nslabs * sizeof(phys_addr_t)));
@@ -447,7 +447,7 @@ void __init swiotlb_exit(void)
} else {
memblock_free_late(__pa(io_tlb_orig_addr),
PAGE_ALIGN(io_tlb_nslabs * sizeof(phys_addr_t)));
- memblock_free_late(__pa(io_tlb_orig_size),
+ memblock_free_late(__pa(io_tlb_alloc_size),
PAGE_ALIGN(io_tlb_nslabs * sizeof(size_t)));
memblock_free_late(__pa(io_tlb_list),
PAGE_ALIGN(io_tlb_nslabs * sizeof(int)));
@@ -639,7 +639,7 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr,
*/
for (i = 0; i < nr_slots(alloc_size + offset); i++) {
io_tlb_orig_addr[index + i] = slot_addr(orig_addr, i);
- io_tlb_orig_size[index+i] = alloc_size - (i << IO_TLB_SHIFT);
+ io_tlb_alloc_size[index+i] = alloc_size - (i << IO_TLB_SHIFT);
}
tlb_addr = slot_addr(io_tlb_start, index) + offset;
if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC) &&
@@ -648,14 +648,14 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr,
return tlb_addr;
}
-static void validate_sync_size_and_truncate(struct device *hwdev, size_t orig_size, size_t *size)
+static void validate_sync_size_and_truncate(struct device *hwdev, size_t alloc_size, size_t *size)
{
- if (*size > orig_size) {
+ if (*size > alloc_size) {
/* Warn and truncate mapping_size */
dev_WARN_ONCE(hwdev, 1,
"Attempt for buffer overflow. Original size: %zu. Mapping size: %zu.\n",
- orig_size, *size);
- *size = orig_size;
+ alloc_size, *size);
+ *size = alloc_size;
}
}
@@ -663,16 +663,17 @@ static void validate_sync_size_and_truncate(struct device *hwdev, size_t orig_si
* tlb_addr is the physical address of the bounce buffer to unmap.
*/
void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
- size_t mapping_size, size_t alloc_size,
- enum dma_data_direction dir, unsigned long attrs)
+ size_t mapping_size, enum dma_data_direction dir,
+ unsigned long attrs)
{
unsigned long flags;
unsigned int offset = swiotlb_align_offset(hwdev, tlb_addr);
- int i, count, nslots = nr_slots(alloc_size + offset);
int index = (tlb_addr - offset - io_tlb_start) >> IO_TLB_SHIFT;
phys_addr_t orig_addr = io_tlb_orig_addr[index];
+ size_t alloc_size = io_tlb_alloc_size[index];
+ int i, count, nslots = nr_slots(alloc_size + offset);
- validate_sync_size_and_truncate(hwdev, io_tlb_orig_size[index], &mapping_size);
+ validate_sync_size_and_truncate(hwdev, alloc_size, &mapping_size);
/*
* First, sync the memory before unmapping the entry
@@ -701,7 +702,7 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
for (i = index + nslots - 1; i >= index; i--) {
io_tlb_list[i] = ++count;
io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
- io_tlb_orig_size[i] = 0;
+ io_tlb_alloc_size[i] = 0;
}
/*
@@ -721,13 +722,13 @@ void swiotlb_tbl_sync_single(struct device *hwdev, phys_addr_t tlb_addr,
enum dma_sync_target target)
{
int index = (tlb_addr - io_tlb_start) >> IO_TLB_SHIFT;
- size_t orig_size = io_tlb_orig_size[index];
+ size_t alloc_size = io_tlb_alloc_size[index];
phys_addr_t orig_addr = io_tlb_orig_addr[index];
if (orig_addr == INVALID_PHYS_ADDR)
return;
- validate_sync_size_and_truncate(hwdev, orig_size, &size);
+ validate_sync_size_and_truncate(hwdev, alloc_size, &size);
switch (target) {
case SYNC_FOR_CPU:
@@ -770,7 +771,7 @@ dma_addr_t swiotlb_map(struct device *dev, phys_addr_t paddr, size_t size,
/* Ensure that the address returned is DMA'ble */
dma_addr = phys_to_dma_unencrypted(dev, swiotlb_addr);
if (unlikely(!dma_capable(dev, dma_addr, size, true))) {
- swiotlb_tbl_unmap_single(dev, swiotlb_addr, size, size, dir,
+ swiotlb_tbl_unmap_single(dev, swiotlb_addr, size, dir,
attrs | DMA_ATTR_SKIP_CPU_SYNC);
dev_WARN_ONCE(dev, 1,
"swiotlb addr %pad+%zu overflow (mask %llx, bus limit %llx).\n",
--
2.29.2
^ permalink raw reply related
* [PATCH 03/14] swiotlb: move orig addr and size validation into swiotlb_bounce
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Move the code to find and validate the original buffer address and size
from the callers into swiotlb_bounce. This means a tiny bit of extra
work in the swiotlb_map path, but avoids code duplication and a leads to
a better code structure.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
kernel/dma/swiotlb.c | 59 +++++++++++++++++---------------------------
1 file changed, 23 insertions(+), 36 deletions(-)
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index 03aa614565e417..a9063092f6f566 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -460,12 +460,25 @@ void __init swiotlb_exit(void)
/*
* Bounce: copy the swiotlb buffer from or back to the original dma location
*/
-static void swiotlb_bounce(phys_addr_t orig_addr, phys_addr_t tlb_addr,
- size_t size, enum dma_data_direction dir)
+static void swiotlb_bounce(struct device *dev, phys_addr_t tlb_addr, size_t size,
+ enum dma_data_direction dir)
{
+ int index = (tlb_addr - io_tlb_start) >> IO_TLB_SHIFT;
+ size_t alloc_size = io_tlb_alloc_size[index];
+ phys_addr_t orig_addr = io_tlb_orig_addr[index];
unsigned long pfn = PFN_DOWN(orig_addr);
unsigned char *vaddr = phys_to_virt(tlb_addr);
+ if (orig_addr == INVALID_PHYS_ADDR)
+ return;
+
+ if (size > alloc_size) {
+ dev_WARN_ONCE(dev, 1,
+ "Buffer overflow detected. Allocation size: %zu. Mapping size: %zu.\n",
+ alloc_size, size);
+ size = alloc_size;
+ }
+
if (PageHighMem(pfn_to_page(pfn))) {
/* The buffer does not have a mapping. Map it in and copy */
unsigned int offset = orig_addr & ~PAGE_MASK;
@@ -644,21 +657,10 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr,
tlb_addr = slot_addr(io_tlb_start, index) + offset;
if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC) &&
(dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL))
- swiotlb_bounce(orig_addr, tlb_addr, mapping_size, DMA_TO_DEVICE);
+ swiotlb_bounce(dev, tlb_addr, mapping_size, DMA_TO_DEVICE);
return tlb_addr;
}
-static void validate_sync_size_and_truncate(struct device *hwdev, size_t alloc_size, size_t *size)
-{
- if (*size > alloc_size) {
- /* Warn and truncate mapping_size */
- dev_WARN_ONCE(hwdev, 1,
- "Attempt for buffer overflow. Original size: %zu. Mapping size: %zu.\n",
- alloc_size, *size);
- *size = alloc_size;
- }
-}
-
/*
* tlb_addr is the physical address of the bounce buffer to unmap.
*/
@@ -669,19 +671,15 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
unsigned long flags;
unsigned int offset = swiotlb_align_offset(hwdev, tlb_addr);
int index = (tlb_addr - offset - io_tlb_start) >> IO_TLB_SHIFT;
- phys_addr_t orig_addr = io_tlb_orig_addr[index];
- size_t alloc_size = io_tlb_alloc_size[index];
- int i, count, nslots = nr_slots(alloc_size + offset);
-
- validate_sync_size_and_truncate(hwdev, alloc_size, &mapping_size);
+ int nslots = nr_slots(io_tlb_alloc_size[index] + offset);
+ int count, i;
/*
* First, sync the memory before unmapping the entry
*/
- if (orig_addr != INVALID_PHYS_ADDR &&
- !(attrs & DMA_ATTR_SKIP_CPU_SYNC) &&
- ((dir == DMA_FROM_DEVICE) || (dir == DMA_BIDIRECTIONAL)))
- swiotlb_bounce(orig_addr, tlb_addr, mapping_size, DMA_FROM_DEVICE);
+ if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC) &&
+ (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL))
+ swiotlb_bounce(hwdev, tlb_addr, mapping_size, DMA_FROM_DEVICE);
/*
* Return the buffer to the free list by setting the corresponding
@@ -721,27 +719,16 @@ void swiotlb_tbl_sync_single(struct device *hwdev, phys_addr_t tlb_addr,
size_t size, enum dma_data_direction dir,
enum dma_sync_target target)
{
- int index = (tlb_addr - io_tlb_start) >> IO_TLB_SHIFT;
- size_t alloc_size = io_tlb_alloc_size[index];
- phys_addr_t orig_addr = io_tlb_orig_addr[index];
-
- if (orig_addr == INVALID_PHYS_ADDR)
- return;
-
- validate_sync_size_and_truncate(hwdev, alloc_size, &size);
-
switch (target) {
case SYNC_FOR_CPU:
if (likely(dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL))
- swiotlb_bounce(orig_addr, tlb_addr,
- size, DMA_FROM_DEVICE);
+ swiotlb_bounce(hwdev, tlb_addr, size, DMA_FROM_DEVICE);
else
BUG_ON(dir != DMA_TO_DEVICE);
break;
case SYNC_FOR_DEVICE:
if (likely(dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL))
- swiotlb_bounce(orig_addr, tlb_addr,
- size, DMA_TO_DEVICE);
+ swiotlb_bounce(hwdev, tlb_addr, size, DMA_TO_DEVICE);
else
BUG_ON(dir != DMA_FROM_DEVICE);
break;
--
2.29.2
^ permalink raw reply related
* [PATCH 04/14] swiotlb: split swiotlb_tbl_sync_single
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Split swiotlb_tbl_sync_single into two separate funtions for the to device
and to cpu synchronization.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/iommu/dma-iommu.c | 12 ++++++------
drivers/xen/swiotlb-xen.c | 4 ++--
include/linux/swiotlb.h | 17 ++++-------------
kernel/dma/direct.c | 8 ++++----
kernel/dma/direct.h | 4 ++--
kernel/dma/swiotlb.c | 34 +++++++++++++++-------------------
6 files changed, 33 insertions(+), 46 deletions(-)
diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
index da2bd8f0885e6e..b57a0e3e21f6c7 100644
--- a/drivers/iommu/dma-iommu.c
+++ b/drivers/iommu/dma-iommu.c
@@ -749,7 +749,7 @@ static void iommu_dma_sync_single_for_cpu(struct device *dev,
arch_sync_dma_for_cpu(phys, size, dir);
if (is_swiotlb_buffer(phys))
- swiotlb_tbl_sync_single(dev, phys, size, dir, SYNC_FOR_CPU);
+ swiotlb_sync_single_for_cpu(dev, phys, size, dir);
}
static void iommu_dma_sync_single_for_device(struct device *dev,
@@ -762,7 +762,7 @@ static void iommu_dma_sync_single_for_device(struct device *dev,
phys = iommu_iova_to_phys(iommu_get_dma_domain(dev), dma_handle);
if (is_swiotlb_buffer(phys))
- swiotlb_tbl_sync_single(dev, phys, size, dir, SYNC_FOR_DEVICE);
+ swiotlb_sync_single_for_device(dev, phys, size, dir);
if (!dev_is_dma_coherent(dev))
arch_sync_dma_for_device(phys, size, dir);
@@ -783,8 +783,8 @@ static void iommu_dma_sync_sg_for_cpu(struct device *dev,
arch_sync_dma_for_cpu(sg_phys(sg), sg->length, dir);
if (is_swiotlb_buffer(sg_phys(sg)))
- swiotlb_tbl_sync_single(dev, sg_phys(sg), sg->length,
- dir, SYNC_FOR_CPU);
+ swiotlb_sync_single_for_cpu(dev, sg_phys(sg),
+ sg->length, dir);
}
}
@@ -800,8 +800,8 @@ static void iommu_dma_sync_sg_for_device(struct device *dev,
for_each_sg(sgl, sg, nelems, i) {
if (is_swiotlb_buffer(sg_phys(sg)))
- swiotlb_tbl_sync_single(dev, sg_phys(sg), sg->length,
- dir, SYNC_FOR_DEVICE);
+ swiotlb_sync_single_for_device(dev, sg_phys(sg),
+ sg->length, dir);
if (!dev_is_dma_coherent(dev))
arch_sync_dma_for_device(sg_phys(sg), sg->length, dir);
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index d47f1b311caac0..4e8a4e14942afd 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -462,7 +462,7 @@ xen_swiotlb_sync_single_for_cpu(struct device *dev, dma_addr_t dma_addr,
}
if (is_xen_swiotlb_buffer(dev, dma_addr))
- swiotlb_tbl_sync_single(dev, paddr, size, dir, SYNC_FOR_CPU);
+ swiotlb_sync_single_for_cpu(dev, paddr, size, dir);
}
static void
@@ -472,7 +472,7 @@ xen_swiotlb_sync_single_for_device(struct device *dev, dma_addr_t dma_addr,
phys_addr_t paddr = xen_dma_to_phys(dev, dma_addr);
if (is_xen_swiotlb_buffer(dev, dma_addr))
- swiotlb_tbl_sync_single(dev, paddr, size, dir, SYNC_FOR_DEVICE);
+ swiotlb_sync_single_for_device(dev, paddr, size, dir);
if (!dev_is_dma_coherent(dev)) {
if (pfn_valid(PFN_DOWN(dma_to_phys(dev, dma_addr))))
diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
index 59f421d041ed9e..0696bdc8072e97 100644
--- a/include/linux/swiotlb.h
+++ b/include/linux/swiotlb.h
@@ -42,14 +42,6 @@ extern int swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs);
extern int swiotlb_late_init_with_default_size(size_t default_size);
extern void __init swiotlb_update_mem_attributes(void);
-/*
- * Enumeration for sync targets
- */
-enum dma_sync_target {
- SYNC_FOR_CPU = 0,
- SYNC_FOR_DEVICE = 1,
-};
-
phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t phys,
size_t mapping_size, size_t alloc_size,
enum dma_data_direction dir, unsigned long attrs);
@@ -60,11 +52,10 @@ extern void swiotlb_tbl_unmap_single(struct device *hwdev,
enum dma_data_direction dir,
unsigned long attrs);
-extern void swiotlb_tbl_sync_single(struct device *hwdev,
- phys_addr_t tlb_addr,
- size_t size, enum dma_data_direction dir,
- enum dma_sync_target target);
-
+void swiotlb_sync_single_for_device(struct device *dev, phys_addr_t tlb_addr,
+ size_t size, enum dma_data_direction dir);
+void swiotlb_sync_single_for_cpu(struct device *dev, phys_addr_t tlb_addr,
+ size_t size, enum dma_data_direction dir);
dma_addr_t swiotlb_map(struct device *dev, phys_addr_t phys,
size_t size, enum dma_data_direction dir, unsigned long attrs);
diff --git a/kernel/dma/direct.c b/kernel/dma/direct.c
index 002268262c9ad8..f737e334705945 100644
--- a/kernel/dma/direct.c
+++ b/kernel/dma/direct.c
@@ -344,8 +344,8 @@ void dma_direct_sync_sg_for_device(struct device *dev,
phys_addr_t paddr = dma_to_phys(dev, sg_dma_address(sg));
if (unlikely(is_swiotlb_buffer(paddr)))
- swiotlb_tbl_sync_single(dev, paddr, sg->length,
- dir, SYNC_FOR_DEVICE);
+ swiotlb_sync_single_for_device(dev, paddr, sg->length,
+ dir);
if (!dev_is_dma_coherent(dev))
arch_sync_dma_for_device(paddr, sg->length,
@@ -370,8 +370,8 @@ void dma_direct_sync_sg_for_cpu(struct device *dev,
arch_sync_dma_for_cpu(paddr, sg->length, dir);
if (unlikely(is_swiotlb_buffer(paddr)))
- swiotlb_tbl_sync_single(dev, paddr, sg->length, dir,
- SYNC_FOR_CPU);
+ swiotlb_sync_single_for_cpu(dev, paddr, sg->length,
+ dir);
if (dir == DMA_FROM_DEVICE)
arch_dma_mark_clean(paddr, sg->length);
diff --git a/kernel/dma/direct.h b/kernel/dma/direct.h
index e1bf721591c0cf..50afc05b6f1dcb 100644
--- a/kernel/dma/direct.h
+++ b/kernel/dma/direct.h
@@ -57,7 +57,7 @@ static inline void dma_direct_sync_single_for_device(struct device *dev,
phys_addr_t paddr = dma_to_phys(dev, addr);
if (unlikely(is_swiotlb_buffer(paddr)))
- swiotlb_tbl_sync_single(dev, paddr, size, dir, SYNC_FOR_DEVICE);
+ swiotlb_sync_single_for_device(dev, paddr, size, dir);
if (!dev_is_dma_coherent(dev))
arch_sync_dma_for_device(paddr, size, dir);
@@ -74,7 +74,7 @@ static inline void dma_direct_sync_single_for_cpu(struct device *dev,
}
if (unlikely(is_swiotlb_buffer(paddr)))
- swiotlb_tbl_sync_single(dev, paddr, size, dir, SYNC_FOR_CPU);
+ swiotlb_sync_single_for_cpu(dev, paddr, size, dir);
if (dir == DMA_FROM_DEVICE)
arch_dma_mark_clean(paddr, size);
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index a9063092f6f566..388d9be35b5795 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -715,26 +715,22 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
spin_unlock_irqrestore(&io_tlb_lock, flags);
}
-void swiotlb_tbl_sync_single(struct device *hwdev, phys_addr_t tlb_addr,
- size_t size, enum dma_data_direction dir,
- enum dma_sync_target target)
+void swiotlb_sync_single_for_device(struct device *dev, phys_addr_t tlb_addr,
+ size_t size, enum dma_data_direction dir)
{
- switch (target) {
- case SYNC_FOR_CPU:
- if (likely(dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL))
- swiotlb_bounce(hwdev, tlb_addr, size, DMA_FROM_DEVICE);
- else
- BUG_ON(dir != DMA_TO_DEVICE);
- break;
- case SYNC_FOR_DEVICE:
- if (likely(dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL))
- swiotlb_bounce(hwdev, tlb_addr, size, DMA_TO_DEVICE);
- else
- BUG_ON(dir != DMA_FROM_DEVICE);
- break;
- default:
- BUG();
- }
+ if (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL)
+ swiotlb_bounce(dev, tlb_addr, size, DMA_TO_DEVICE);
+ else
+ BUG_ON(dir != DMA_FROM_DEVICE);
+}
+
+void swiotlb_sync_single_for_cpu(struct device *dev, phys_addr_t tlb_addr,
+ size_t size, enum dma_data_direction dir)
+{
+ if (dir == DMA_FROM_DEVICE || dir == DMA_BIDIRECTIONAL)
+ swiotlb_bounce(dev, tlb_addr, size, DMA_FROM_DEVICE);
+ else
+ BUG_ON(dir != DMA_TO_DEVICE);
}
/*
--
2.29.2
^ permalink raw reply related
* [PATCH 05/14] xen-swiotlb: use is_swiotlb_buffer in is_xen_swiotlb_buffer
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Use the is_swiotlb_buffer to check if a physical address is
a swiotlb buffer. This works because xen-swiotlb does use the
same buffer as the main swiotlb code, and xen_io_tlb_{start,end}
are just the addresses for it that went through phys_to_virt.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
drivers/xen/swiotlb-xen.c | 6 ++----
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 4e8a4e14942afd..bffb35993c9d5f 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -111,10 +111,8 @@ static int is_xen_swiotlb_buffer(struct device *dev, dma_addr_t dma_addr)
* have the same virtual address as another address
* in our domain. Therefore _only_ check address within our domain.
*/
- if (pfn_valid(PFN_DOWN(paddr))) {
- return paddr >= virt_to_phys(xen_io_tlb_start) &&
- paddr < virt_to_phys(xen_io_tlb_end);
- }
+ if (pfn_valid(PFN_DOWN(paddr)))
+ return is_swiotlb_buffer(paddr);
return 0;
}
--
2.29.2
^ permalink raw reply related
* [PATCH 06/14] xen-swiotlb: use io_tlb_end in xen_swiotlb_dma_supported
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Use the existing variable that holds the physical address for
xen_io_tlb_end to simplify xen_swiotlb_dma_supported a bit, and remove
the otherwise unused xen_io_tlb_end variable and the xen_virt_to_bus
helper.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
---
drivers/xen/swiotlb-xen.c | 10 ++--------
1 file changed, 2 insertions(+), 8 deletions(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index bffb35993c9d5f..e99f0614dcb979 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -46,7 +46,7 @@
* API.
*/
-static char *xen_io_tlb_start, *xen_io_tlb_end;
+static char *xen_io_tlb_start;
static unsigned long xen_io_tlb_nslabs;
/*
* Quick lookup value of the bus address of the IOTLB.
@@ -82,11 +82,6 @@ static inline phys_addr_t xen_dma_to_phys(struct device *dev,
return xen_bus_to_phys(dev, dma_to_phys(dev, dma_addr));
}
-static inline dma_addr_t xen_virt_to_bus(struct device *dev, void *address)
-{
- return xen_phys_to_dma(dev, virt_to_phys(address));
-}
-
static inline int range_straddles_page_boundary(phys_addr_t p, size_t size)
{
unsigned long next_bfn, xen_pfn = XEN_PFN_DOWN(p);
@@ -250,7 +245,6 @@ int __ref xen_swiotlb_init(int verbose, bool early)
rc = swiotlb_late_init_with_tbl(xen_io_tlb_start, xen_io_tlb_nslabs);
end:
- xen_io_tlb_end = xen_io_tlb_start + bytes;
if (!rc)
swiotlb_set_max_segment(PAGE_SIZE);
@@ -558,7 +552,7 @@ xen_swiotlb_sync_sg_for_device(struct device *dev, struct scatterlist *sgl,
static int
xen_swiotlb_dma_supported(struct device *hwdev, u64 mask)
{
- return xen_virt_to_bus(hwdev, xen_io_tlb_end - 1) <= mask;
+ return xen_phys_to_dma(hwdev, io_tlb_end - 1) <= mask;
}
const struct dma_map_ops xen_swiotlb_dma_ops = {
--
2.29.2
^ permalink raw reply related
* [PATCH 07/14] xen-swiotlb: remove xen_set_nslabs
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
The xen_set_nslabs function is a little weird, as it has just one
caller, that caller passes a global variable as the argument,
which is then overriden in the function and a derivative of it
returned. Just add a cpp symbol for the default size using a readable
constant and open code the remaining three lines in the caller.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/xen/swiotlb-xen.c | 19 +++++++------------
1 file changed, 7 insertions(+), 12 deletions(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index e99f0614dcb979..5352655432e724 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -138,16 +138,6 @@ xen_swiotlb_fixup(void *buf, size_t size, unsigned long nslabs)
} while (i < nslabs);
return 0;
}
-static unsigned long xen_set_nslabs(unsigned long nr_tbl)
-{
- if (!nr_tbl) {
- xen_io_tlb_nslabs = (64 * 1024 * 1024 >> IO_TLB_SHIFT);
- xen_io_tlb_nslabs = ALIGN(xen_io_tlb_nslabs, IO_TLB_SEGSIZE);
- } else
- xen_io_tlb_nslabs = nr_tbl;
-
- return xen_io_tlb_nslabs << IO_TLB_SHIFT;
-}
enum xen_swiotlb_err {
XEN_SWIOTLB_UNKNOWN = 0,
@@ -170,6 +160,9 @@ static const char *xen_swiotlb_error(enum xen_swiotlb_err err)
}
return "";
}
+
+#define DEFAULT_NSLABS ALIGN(SZ_64M >> IO_TLB_SHIFT, IO_TLB_SEGSIZE)
+
int __ref xen_swiotlb_init(int verbose, bool early)
{
unsigned long bytes, order;
@@ -179,8 +172,10 @@ int __ref xen_swiotlb_init(int verbose, bool early)
xen_io_tlb_nslabs = swiotlb_nr_tbl();
retry:
- bytes = xen_set_nslabs(xen_io_tlb_nslabs);
- order = get_order(xen_io_tlb_nslabs << IO_TLB_SHIFT);
+ if (!xen_io_tlb_nslabs)
+ xen_io_tlb_nslabs = DEFAULT_NSLABS;
+ bytes = xen_io_tlb_nslabs << IO_TLB_SHIFT;
+ order = get_order(bytes);
/*
* IO TLB memory already allocated. Just use it.
--
2.29.2
^ permalink raw reply related
* [PATCH 08/14] xen-swiotlb: remove xen_io_tlb_start and xen_io_tlb_nslabs
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
The xen_io_tlb_start and xen_io_tlb_nslabs variables ar now only used in
xen_swiotlb_init, so replace them with local variables.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/xen/swiotlb-xen.c | 57 +++++++++++++++++----------------------
1 file changed, 25 insertions(+), 32 deletions(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 5352655432e724..1a31ddf7139799 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -40,14 +40,7 @@
#include <trace/events/swiotlb.h>
#define MAX_DMA_BITS 32
-/*
- * Used to do a quick range check in swiotlb_tbl_unmap_single and
- * swiotlb_tbl_sync_single_*, to see if the memory was in fact allocated by this
- * API.
- */
-static char *xen_io_tlb_start;
-static unsigned long xen_io_tlb_nslabs;
/*
* Quick lookup value of the bus address of the IOTLB.
*/
@@ -169,75 +162,75 @@ int __ref xen_swiotlb_init(int verbose, bool early)
int rc = -ENOMEM;
enum xen_swiotlb_err m_ret = XEN_SWIOTLB_UNKNOWN;
unsigned int repeat = 3;
+ char *start;
+ unsigned long nslabs;
- xen_io_tlb_nslabs = swiotlb_nr_tbl();
+ nslabs = swiotlb_nr_tbl();
retry:
- if (!xen_io_tlb_nslabs)
- xen_io_tlb_nslabs = DEFAULT_NSLABS;
- bytes = xen_io_tlb_nslabs << IO_TLB_SHIFT;
+ if (!nslabs)
+ nslabs = DEFAULT_NSLABS;
+ bytes = nslabs << IO_TLB_SHIFT;
order = get_order(bytes);
/*
* IO TLB memory already allocated. Just use it.
*/
- if (io_tlb_start != 0) {
- xen_io_tlb_start = phys_to_virt(io_tlb_start);
+ if (io_tlb_start != 0)
goto end;
- }
/*
* Get IO TLB memory from any location.
*/
if (early) {
- xen_io_tlb_start = memblock_alloc(PAGE_ALIGN(bytes),
+ start = memblock_alloc(PAGE_ALIGN(bytes),
PAGE_SIZE);
- if (!xen_io_tlb_start)
+ if (!start)
panic("%s: Failed to allocate %lu bytes align=0x%lx\n",
__func__, PAGE_ALIGN(bytes), PAGE_SIZE);
} else {
#define SLABS_PER_PAGE (1 << (PAGE_SHIFT - IO_TLB_SHIFT))
#define IO_TLB_MIN_SLABS ((1<<20) >> IO_TLB_SHIFT)
while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) {
- xen_io_tlb_start = (void *)xen_get_swiotlb_free_pages(order);
- if (xen_io_tlb_start)
+ start = (void *)xen_get_swiotlb_free_pages(order);
+ if (start)
break;
order--;
}
if (order != get_order(bytes)) {
pr_warn("Warning: only able to allocate %ld MB for software IO TLB\n",
(PAGE_SIZE << order) >> 20);
- xen_io_tlb_nslabs = SLABS_PER_PAGE << order;
- bytes = xen_io_tlb_nslabs << IO_TLB_SHIFT;
+ nslabs = SLABS_PER_PAGE << order;
+ bytes = nslabs << IO_TLB_SHIFT;
}
}
- if (!xen_io_tlb_start) {
+ if (!start) {
m_ret = XEN_SWIOTLB_ENOMEM;
goto error;
}
/*
* And replace that memory with pages under 4GB.
*/
- rc = xen_swiotlb_fixup(xen_io_tlb_start,
+ rc = xen_swiotlb_fixup(start,
bytes,
- xen_io_tlb_nslabs);
+ nslabs);
if (rc) {
if (early)
- memblock_free(__pa(xen_io_tlb_start),
+ memblock_free(__pa(start),
PAGE_ALIGN(bytes));
else {
- free_pages((unsigned long)xen_io_tlb_start, order);
- xen_io_tlb_start = NULL;
+ free_pages((unsigned long)start, order);
+ start = NULL;
}
m_ret = XEN_SWIOTLB_EFIXUP;
goto error;
}
if (early) {
- if (swiotlb_init_with_tbl(xen_io_tlb_start, xen_io_tlb_nslabs,
+ if (swiotlb_init_with_tbl(start, nslabs,
verbose))
panic("Cannot allocate SWIOTLB buffer");
rc = 0;
} else
- rc = swiotlb_late_init_with_tbl(xen_io_tlb_start, xen_io_tlb_nslabs);
+ rc = swiotlb_late_init_with_tbl(start, nslabs);
end:
if (!rc)
@@ -246,17 +239,17 @@ int __ref xen_swiotlb_init(int verbose, bool early)
return rc;
error:
if (repeat--) {
- xen_io_tlb_nslabs = max(1024UL, /* Min is 2MB */
- (xen_io_tlb_nslabs >> 1));
+ nslabs = max(1024UL, /* Min is 2MB */
+ (nslabs >> 1));
pr_info("Lowering to %luMB\n",
- (xen_io_tlb_nslabs << IO_TLB_SHIFT) >> 20);
+ (nslabs << IO_TLB_SHIFT) >> 20);
goto retry;
}
pr_err("%s (rc:%d)\n", xen_swiotlb_error(m_ret), rc);
if (early)
panic("%s (rc:%d)", xen_swiotlb_error(m_ret), rc);
else
- free_pages((unsigned long)xen_io_tlb_start, order);
+ free_pages((unsigned long)start, order);
return rc;
}
--
2.29.2
^ permalink raw reply related
* [PATCH 09/14] swiotlb: lift the double initialization protection from xen-swiotlb
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Lift the double initialization protection from xen-swiotlb to the core
code to avoid exposing too many swiotlb internals. Also upgrade the
check to a warning as it should not happen.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/xen/swiotlb-xen.c | 7 -------
kernel/dma/swiotlb.c | 8 ++++++++
2 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 1a31ddf7139799..060eeb056486f5 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -172,12 +172,6 @@ int __ref xen_swiotlb_init(int verbose, bool early)
bytes = nslabs << IO_TLB_SHIFT;
order = get_order(bytes);
- /*
- * IO TLB memory already allocated. Just use it.
- */
- if (io_tlb_start != 0)
- goto end;
-
/*
* Get IO TLB memory from any location.
*/
@@ -232,7 +226,6 @@ int __ref xen_swiotlb_init(int verbose, bool early)
} else
rc = swiotlb_late_init_with_tbl(start, nslabs);
-end:
if (!rc)
swiotlb_set_max_segment(PAGE_SIZE);
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index 388d9be35b5795..ebe7c123e27e52 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -229,6 +229,10 @@ int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose)
unsigned long i, bytes;
size_t alloc_size;
+ /* protect against double initialization */
+ if (WARN_ON_ONCE(io_tlb_start))
+ return -ENOMEM;
+
bytes = nslabs << IO_TLB_SHIFT;
io_tlb_nslabs = nslabs;
@@ -367,6 +371,10 @@ swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
{
unsigned long i, bytes;
+ /* protect against double initialization */
+ if (WARN_ON_ONCE(io_tlb_start))
+ return -ENOMEM;
+
bytes = nslabs << IO_TLB_SHIFT;
io_tlb_nslabs = nslabs;
--
2.29.2
^ permalink raw reply related
* [PATCH 10/14] xen-swiotlb: split xen_swiotlb_init
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Split xen_swiotlb_init into a normal an an early case. That makes both
much simpler and more readable, and also allows marking the early
code as __init and x86-only.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
arch/arm/xen/mm.c | 2 +-
arch/x86/xen/pci-swiotlb-xen.c | 4 +-
drivers/xen/swiotlb-xen.c | 124 +++++++++++++++++++--------------
include/xen/swiotlb-xen.h | 3 +-
4 files changed, 75 insertions(+), 58 deletions(-)
diff --git a/arch/arm/xen/mm.c b/arch/arm/xen/mm.c
index 467fa225c3d0ed..aae950cd053fea 100644
--- a/arch/arm/xen/mm.c
+++ b/arch/arm/xen/mm.c
@@ -140,7 +140,7 @@ static int __init xen_mm_init(void)
struct gnttab_cache_flush cflush;
if (!xen_initial_domain())
return 0;
- xen_swiotlb_init(1, false);
+ xen_swiotlb_init();
cflush.op = 0;
cflush.a.dev_bus_addr = 0;
diff --git a/arch/x86/xen/pci-swiotlb-xen.c b/arch/x86/xen/pci-swiotlb-xen.c
index 19ae3e4fe4e98e..54f9aa7e845739 100644
--- a/arch/x86/xen/pci-swiotlb-xen.c
+++ b/arch/x86/xen/pci-swiotlb-xen.c
@@ -59,7 +59,7 @@ int __init pci_xen_swiotlb_detect(void)
void __init pci_xen_swiotlb_init(void)
{
if (xen_swiotlb) {
- xen_swiotlb_init(1, true /* early */);
+ xen_swiotlb_init_early();
dma_ops = &xen_swiotlb_dma_ops;
#ifdef CONFIG_PCI
@@ -76,7 +76,7 @@ int pci_xen_swiotlb_init_late(void)
if (xen_swiotlb)
return 0;
- rc = xen_swiotlb_init(1, false /* late */);
+ rc = xen_swiotlb_init();
if (rc)
return rc;
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 060eeb056486f5..00adeb95ebb9df 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -156,96 +156,112 @@ static const char *xen_swiotlb_error(enum xen_swiotlb_err err)
#define DEFAULT_NSLABS ALIGN(SZ_64M >> IO_TLB_SHIFT, IO_TLB_SEGSIZE)
-int __ref xen_swiotlb_init(int verbose, bool early)
+int __ref xen_swiotlb_init(void)
{
- unsigned long bytes, order;
- int rc = -ENOMEM;
enum xen_swiotlb_err m_ret = XEN_SWIOTLB_UNKNOWN;
+ unsigned long nslabs, bytes, order;
unsigned int repeat = 3;
+ int rc = -ENOMEM;
char *start;
- unsigned long nslabs;
nslabs = swiotlb_nr_tbl();
-retry:
if (!nslabs)
nslabs = DEFAULT_NSLABS;
+retry:
+ m_ret = XEN_SWIOTLB_ENOMEM;
bytes = nslabs << IO_TLB_SHIFT;
order = get_order(bytes);
/*
* Get IO TLB memory from any location.
*/
- if (early) {
- start = memblock_alloc(PAGE_ALIGN(bytes),
- PAGE_SIZE);
- if (!start)
- panic("%s: Failed to allocate %lu bytes align=0x%lx\n",
- __func__, PAGE_ALIGN(bytes), PAGE_SIZE);
- } else {
#define SLABS_PER_PAGE (1 << (PAGE_SHIFT - IO_TLB_SHIFT))
#define IO_TLB_MIN_SLABS ((1<<20) >> IO_TLB_SHIFT)
- while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) {
- start = (void *)xen_get_swiotlb_free_pages(order);
- if (start)
- break;
- order--;
- }
- if (order != get_order(bytes)) {
- pr_warn("Warning: only able to allocate %ld MB for software IO TLB\n",
- (PAGE_SIZE << order) >> 20);
- nslabs = SLABS_PER_PAGE << order;
- bytes = nslabs << IO_TLB_SHIFT;
- }
+ while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) {
+ start = (void *)xen_get_swiotlb_free_pages(order);
+ if (start)
+ break;
+ order--;
}
- if (!start) {
- m_ret = XEN_SWIOTLB_ENOMEM;
+ if (!start)
goto error;
+ if (order != get_order(bytes)) {
+ pr_warn("Warning: only able to allocate %ld MB for software IO TLB\n",
+ (PAGE_SIZE << order) >> 20);
+ nslabs = SLABS_PER_PAGE << order;
+ bytes = nslabs << IO_TLB_SHIFT;
}
+
/*
* And replace that memory with pages under 4GB.
*/
- rc = xen_swiotlb_fixup(start,
- bytes,
- nslabs);
+ rc = xen_swiotlb_fixup(start, bytes, nslabs);
if (rc) {
- if (early)
- memblock_free(__pa(start),
- PAGE_ALIGN(bytes));
- else {
- free_pages((unsigned long)start, order);
- start = NULL;
- }
+ free_pages((unsigned long)start, order);
m_ret = XEN_SWIOTLB_EFIXUP;
goto error;
}
- if (early) {
- if (swiotlb_init_with_tbl(start, nslabs,
- verbose))
- panic("Cannot allocate SWIOTLB buffer");
- rc = 0;
- } else
- rc = swiotlb_late_init_with_tbl(start, nslabs);
-
- if (!rc)
- swiotlb_set_max_segment(PAGE_SIZE);
-
- return rc;
+ rc = swiotlb_late_init_with_tbl(start, nslabs);
+ if (rc)
+ return rc;
+ swiotlb_set_max_segment(PAGE_SIZE);
+ return 0;
error:
if (repeat--) {
- nslabs = max(1024UL, /* Min is 2MB */
- (nslabs >> 1));
+ /* Min is 2MB */
+ nslabs = max(1024UL, (nslabs >> 1));
pr_info("Lowering to %luMB\n",
(nslabs << IO_TLB_SHIFT) >> 20);
goto retry;
}
pr_err("%s (rc:%d)\n", xen_swiotlb_error(m_ret), rc);
- if (early)
- panic("%s (rc:%d)", xen_swiotlb_error(m_ret), rc);
- else
- free_pages((unsigned long)start, order);
+ free_pages((unsigned long)start, order);
return rc;
}
+#ifdef CONFIG_X86
+void __init xen_swiotlb_init_early(void)
+{
+ unsigned long nslabs, bytes;
+ unsigned int repeat = 3;
+ char *start;
+ int rc;
+
+ nslabs = swiotlb_nr_tbl();
+ if (!nslabs)
+ nslabs = DEFAULT_NSLABS;
+retry:
+ /*
+ * Get IO TLB memory from any location.
+ */
+ bytes = nslabs << IO_TLB_SHIFT;
+ start = memblock_alloc(PAGE_ALIGN(bytes), PAGE_SIZE);
+ if (!start)
+ panic("%s: Failed to allocate %lu bytes align=0x%lx\n",
+ __func__, PAGE_ALIGN(bytes), PAGE_SIZE);
+
+ /*
+ * And replace that memory with pages under 4GB.
+ */
+ rc = xen_swiotlb_fixup(start, bytes, nslabs);
+ if (rc) {
+ memblock_free(__pa(start), PAGE_ALIGN(bytes));
+ if (repeat--) {
+ /* Min is 2MB */
+ nslabs = max(1024UL, (nslabs >> 1));
+ pr_info("Lowering to %luMB\n",
+ (nslabs << IO_TLB_SHIFT) >> 20);
+ goto retry;
+ }
+ panic("%s (rc:%d)", xen_swiotlb_error(XEN_SWIOTLB_EFIXUP), rc);
+ }
+
+ if (swiotlb_init_with_tbl(start, nslabs, false))
+ panic("Cannot allocate SWIOTLB buffer");
+ swiotlb_set_max_segment(PAGE_SIZE);
+}
+#endif /* CONFIG_X86 */
+
static void *
xen_swiotlb_alloc_coherent(struct device *hwdev, size_t size,
dma_addr_t *dma_handle, gfp_t flags,
diff --git a/include/xen/swiotlb-xen.h b/include/xen/swiotlb-xen.h
index d5eaf9d682b804..6206b1ec99168a 100644
--- a/include/xen/swiotlb-xen.h
+++ b/include/xen/swiotlb-xen.h
@@ -9,7 +9,8 @@ void xen_dma_sync_for_cpu(struct device *dev, dma_addr_t handle,
void xen_dma_sync_for_device(struct device *dev, dma_addr_t handle,
size_t size, enum dma_data_direction dir);
-extern int xen_swiotlb_init(int verbose, bool early);
+int xen_swiotlb_init(void);
+void __init xen_swiotlb_init_early(void);
extern const struct dma_map_ops xen_swiotlb_dma_ops;
#endif /* __LINUX_SWIOTLB_XEN_H */
--
2.29.2
^ permalink raw reply related
* [PATCH 11/14] xen-swiotlb: remove the unused size argument from xen_swiotlb_fixup
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/xen/swiotlb-xen.c | 7 +++----
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 00adeb95ebb9df..4ecfce2c6f7263 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -104,8 +104,7 @@ static int is_xen_swiotlb_buffer(struct device *dev, dma_addr_t dma_addr)
return 0;
}
-static int
-xen_swiotlb_fixup(void *buf, size_t size, unsigned long nslabs)
+static int xen_swiotlb_fixup(void *buf, unsigned long nslabs)
{
int i, rc;
int dma_bits;
@@ -195,7 +194,7 @@ int __ref xen_swiotlb_init(void)
/*
* And replace that memory with pages under 4GB.
*/
- rc = xen_swiotlb_fixup(start, bytes, nslabs);
+ rc = xen_swiotlb_fixup(start, nslabs);
if (rc) {
free_pages((unsigned long)start, order);
m_ret = XEN_SWIOTLB_EFIXUP;
@@ -243,7 +242,7 @@ void __init xen_swiotlb_init_early(void)
/*
* And replace that memory with pages under 4GB.
*/
- rc = xen_swiotlb_fixup(start, bytes, nslabs);
+ rc = xen_swiotlb_fixup(start, nslabs);
if (rc) {
memblock_free(__pa(start), PAGE_ALIGN(bytes));
if (repeat--) {
--
2.29.2
^ permalink raw reply related
* [PATCH 12/14] swiotlb: move global variables into a new io_tlb_mem structure
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
From: Claire Chang <tientzu@chromium.org>
Added a new struct, io_tlb_mem, as the IO TLB memory pool descriptor and
moved relevant global variables into that struct.
This will be useful later to allow for restricted DMA pool.
Signed-off-by: Claire Chang <tientzu@chromium.org>
[hch: rebased]
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/xen/swiotlb-xen.c | 2 +-
include/linux/swiotlb.h | 43 ++++-
kernel/dma/swiotlb.c | 354 +++++++++++++++++---------------------
3 files changed, 203 insertions(+), 196 deletions(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 4ecfce2c6f7263..5329ad54a5f34e 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -548,7 +548,7 @@ xen_swiotlb_sync_sg_for_device(struct device *dev, struct scatterlist *sgl,
static int
xen_swiotlb_dma_supported(struct device *hwdev, u64 mask)
{
- return xen_phys_to_dma(hwdev, io_tlb_end - 1) <= mask;
+ return xen_phys_to_dma(hwdev, io_tlb_default_mem.end - 1) <= mask;
}
const struct dma_map_ops xen_swiotlb_dma_ops = {
diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
index 0696bdc8072e97..5ec5378b17c333 100644
--- a/include/linux/swiotlb.h
+++ b/include/linux/swiotlb.h
@@ -6,6 +6,7 @@
#include <linux/init.h>
#include <linux/types.h>
#include <linux/limits.h>
+#include <linux/spinlock.h>
struct device;
struct page;
@@ -61,11 +62,49 @@ dma_addr_t swiotlb_map(struct device *dev, phys_addr_t phys,
#ifdef CONFIG_SWIOTLB
extern enum swiotlb_force swiotlb_force;
-extern phys_addr_t io_tlb_start, io_tlb_end;
+
+/**
+ * struct io_tlb_mem - IO TLB Memory Pool Descriptor
+ *
+ * @start: The start address of the swiotlb memory pool. Used to do a quick
+ * range check to see if the memory was in fact allocated by this
+ * API.
+ * @end: The end address of the swiotlb memory pool. Used to do a quick
+ * range check to see if the memory was in fact allocated by this
+ * API.
+ * @nslabs: The number of IO TLB blocks (in groups of 64) between @start and
+ * @end. This is command line adjustable via setup_io_tlb_npages.
+ * @used: The number of used IO TLB block.
+ * @list: The free list describing the number of free entries available
+ * from each index.
+ * @index: The index to start searching in the next round.
+ * @orig_addr: The original address corresponding to a mapped entry.
+ * @alloc_size: Size of the allocated buffer.
+ * @lock: The lock to protect the above data structures in the map and
+ * unmap calls.
+ * @debugfs: The dentry to debugfs.
+ * @late_alloc: %true if allocated using the page allocator
+ */
+struct io_tlb_mem {
+ phys_addr_t start;
+ phys_addr_t end;
+ unsigned long nslabs;
+ unsigned long used;
+ unsigned int *list;
+ unsigned int index;
+ phys_addr_t *orig_addr;
+ size_t *alloc_size;
+ spinlock_t lock;
+ struct dentry *debugfs;
+ bool late_alloc;
+};
+extern struct io_tlb_mem io_tlb_default_mem;
static inline bool is_swiotlb_buffer(phys_addr_t paddr)
{
- return paddr >= io_tlb_start && paddr < io_tlb_end;
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
+
+ return paddr >= mem->start && paddr < mem->end;
}
void __init swiotlb_exit(void);
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index ebe7c123e27e52..6aa84fa3b1467e 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -59,32 +59,11 @@
*/
#define IO_TLB_MIN_SLABS ((1<<20) >> IO_TLB_SHIFT)
-enum swiotlb_force swiotlb_force;
-
-/*
- * Used to do a quick range check in swiotlb_tbl_unmap_single and
- * swiotlb_tbl_sync_single_*, to see if the memory was in fact allocated by this
- * API.
- */
-phys_addr_t io_tlb_start, io_tlb_end;
-
-/*
- * The number of IO TLB blocks (in groups of 64) between io_tlb_start and
- * io_tlb_end. This is command line adjustable via setup_io_tlb_npages.
- */
-static unsigned long io_tlb_nslabs;
+#define INVALID_PHYS_ADDR (~(phys_addr_t)0)
-/*
- * The number of used IO TLB block
- */
-static unsigned long io_tlb_used;
+enum swiotlb_force swiotlb_force;
-/*
- * This is a free list describing the number of free entries available from
- * each index
- */
-static unsigned int *io_tlb_list;
-static unsigned int io_tlb_index;
+struct io_tlb_mem io_tlb_default_mem;
/*
* Max segment that we can provide which (if pages are contingous) will
@@ -92,32 +71,15 @@ static unsigned int io_tlb_index;
*/
static unsigned int max_segment;
-/*
- * We need to save away the original address corresponding to a mapped entry
- * for the sync operations.
- */
-#define INVALID_PHYS_ADDR (~(phys_addr_t)0)
-static phys_addr_t *io_tlb_orig_addr;
-
-/*
- * The mapped buffer's size should be validated during a sync operation.
- */
-static size_t *io_tlb_alloc_size;
-
-/*
- * Protect the above data structures in the map and unmap calls
- */
-static DEFINE_SPINLOCK(io_tlb_lock);
-
-static int late_alloc;
-
static int __init
setup_io_tlb_npages(char *str)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
+
if (isdigit(*str)) {
- io_tlb_nslabs = simple_strtoul(str, &str, 0);
+ mem->nslabs = simple_strtoul(str, &str, 0);
/* avoid tail segment of size < IO_TLB_SEGSIZE */
- io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+ mem->nslabs = ALIGN(mem->nslabs, IO_TLB_SEGSIZE);
}
if (*str == ',')
++str;
@@ -125,7 +87,7 @@ setup_io_tlb_npages(char *str)
swiotlb_force = SWIOTLB_FORCE;
} else if (!strcmp(str, "noforce")) {
swiotlb_force = SWIOTLB_NO_FORCE;
- io_tlb_nslabs = 1;
+ mem->nslabs = 1;
}
return 0;
@@ -136,7 +98,7 @@ static bool no_iotlb_memory;
unsigned long swiotlb_nr_tbl(void)
{
- return unlikely(no_iotlb_memory) ? 0 : io_tlb_nslabs;
+ return unlikely(no_iotlb_memory) ? 0 : io_tlb_default_mem.nslabs;
}
EXPORT_SYMBOL_GPL(swiotlb_nr_tbl);
@@ -158,13 +120,14 @@ unsigned long swiotlb_size_or_default(void)
{
unsigned long size;
- size = io_tlb_nslabs << IO_TLB_SHIFT;
+ size = io_tlb_default_mem.nslabs << IO_TLB_SHIFT;
return size ? size : (IO_TLB_DEFAULT_SIZE);
}
void __init swiotlb_adjust_size(unsigned long new_size)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
unsigned long size;
/*
@@ -172,10 +135,10 @@ void __init swiotlb_adjust_size(unsigned long new_size)
* architectures such as those supporting memory encryption to
* adjust/expand SWIOTLB size for their use.
*/
- if (!io_tlb_nslabs) {
+ if (!mem->nslabs) {
size = ALIGN(new_size, IO_TLB_SIZE);
- io_tlb_nslabs = size >> IO_TLB_SHIFT;
- io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+ mem->nslabs = size >> IO_TLB_SHIFT;
+ mem->nslabs = ALIGN(mem->nslabs, IO_TLB_SEGSIZE);
pr_info("SWIOTLB bounce buffer size adjusted to %luMB", size >> 20);
}
@@ -183,14 +146,15 @@ void __init swiotlb_adjust_size(unsigned long new_size)
void swiotlb_print_info(void)
{
- unsigned long bytes = io_tlb_nslabs << IO_TLB_SHIFT;
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
+ unsigned long bytes = mem->nslabs << IO_TLB_SHIFT;
if (no_iotlb_memory) {
pr_warn("No low mem\n");
return;
}
- pr_info("mapped [mem %pa-%pa] (%luMB)\n", &io_tlb_start, &io_tlb_end,
+ pr_info("mapped [mem %pa-%pa] (%luMB)\n", &mem->start, &mem->end,
bytes >> 20);
}
@@ -212,68 +176,65 @@ static inline unsigned long nr_slots(u64 val)
*/
void __init swiotlb_update_mem_attributes(void)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
void *vaddr;
unsigned long bytes;
- if (no_iotlb_memory || late_alloc)
+ if (no_iotlb_memory || mem->late_alloc)
return;
- vaddr = phys_to_virt(io_tlb_start);
- bytes = PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT);
+ vaddr = phys_to_virt(mem->start);
+ bytes = PAGE_ALIGN(mem->nslabs << IO_TLB_SHIFT);
set_memory_decrypted((unsigned long)vaddr, bytes >> PAGE_SHIFT);
memset(vaddr, 0, bytes);
}
int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
unsigned long i, bytes;
size_t alloc_size;
/* protect against double initialization */
- if (WARN_ON_ONCE(io_tlb_start))
+ if (WARN_ON_ONCE(mem->start))
return -ENOMEM;
bytes = nslabs << IO_TLB_SHIFT;
- io_tlb_nslabs = nslabs;
- io_tlb_start = __pa(tlb);
- io_tlb_end = io_tlb_start + bytes;
+ mem->nslabs = nslabs;
+ mem->start = __pa(tlb);
+ mem->end = mem->start + bytes;
+ mem->index = 0;
+ spin_lock_init(&mem->lock);
/*
* Allocate and initialize the free list array. This array is used
* to find contiguous free memory regions of size up to IO_TLB_SEGSIZE
- * between io_tlb_start and io_tlb_end.
+ * between mem->start and mem->end.
*/
- alloc_size = PAGE_ALIGN(io_tlb_nslabs * sizeof(int));
- io_tlb_list = memblock_alloc(alloc_size, PAGE_SIZE);
- if (!io_tlb_list)
+ alloc_size = PAGE_ALIGN(mem->nslabs * sizeof(int));
+ mem->list = memblock_alloc(alloc_size, PAGE_SIZE);
+ if (!mem->list)
panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
__func__, alloc_size, PAGE_SIZE);
- alloc_size = PAGE_ALIGN(io_tlb_nslabs * sizeof(phys_addr_t));
- io_tlb_orig_addr = memblock_alloc(alloc_size, PAGE_SIZE);
- if (!io_tlb_orig_addr)
+ alloc_size = PAGE_ALIGN(mem->nslabs * sizeof(phys_addr_t));
+ mem->orig_addr = memblock_alloc(alloc_size, PAGE_SIZE);
+ if (!mem->orig_addr)
panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
__func__, alloc_size, PAGE_SIZE);
- alloc_size = PAGE_ALIGN(io_tlb_nslabs * sizeof(size_t));
- io_tlb_alloc_size = memblock_alloc(alloc_size, PAGE_SIZE);
- if (!io_tlb_alloc_size)
- panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
- __func__, alloc_size, PAGE_SIZE);
-
- for (i = 0; i < io_tlb_nslabs; i++) {
- io_tlb_list[i] = IO_TLB_SEGSIZE - io_tlb_offset(i);
- io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
- io_tlb_alloc_size[i] = 0;
+ for (i = 0; i < mem->nslabs; i++) {
+ mem->list[i] = IO_TLB_SEGSIZE - io_tlb_offset(i);
+ mem->orig_addr[i] = INVALID_PHYS_ADDR;
+ mem->alloc_size[i] = 0;
}
- io_tlb_index = 0;
no_iotlb_memory = false;
if (verbose)
swiotlb_print_info();
- swiotlb_set_max_segment(io_tlb_nslabs << IO_TLB_SHIFT);
+ swiotlb_set_max_segment(mem->nslabs << IO_TLB_SHIFT);
return 0;
}
@@ -284,26 +245,27 @@ int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose)
void __init
swiotlb_init(int verbose)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
size_t default_size = IO_TLB_DEFAULT_SIZE;
unsigned char *vstart;
unsigned long bytes;
- if (!io_tlb_nslabs) {
- io_tlb_nslabs = (default_size >> IO_TLB_SHIFT);
- io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+ if (!mem->nslabs) {
+ mem->nslabs = (default_size >> IO_TLB_SHIFT);
+ mem->nslabs = ALIGN(mem->nslabs, IO_TLB_SEGSIZE);
}
- bytes = io_tlb_nslabs << IO_TLB_SHIFT;
+ bytes = mem->nslabs << IO_TLB_SHIFT;
/* Get IO TLB memory from the low pages */
vstart = memblock_alloc_low(PAGE_ALIGN(bytes), PAGE_SIZE);
- if (vstart && !swiotlb_init_with_tbl(vstart, io_tlb_nslabs, verbose))
+ if (vstart && !swiotlb_init_with_tbl(vstart, mem->nslabs, verbose))
return;
- if (io_tlb_start) {
- memblock_free_early(io_tlb_start,
- PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT));
- io_tlb_start = 0;
+ if (mem->start) {
+ memblock_free_early(mem->start,
+ PAGE_ALIGN(mem->nslabs << IO_TLB_SHIFT));
+ mem->start = 0;
}
pr_warn("Cannot allocate buffer");
no_iotlb_memory = true;
@@ -317,22 +279,23 @@ swiotlb_init(int verbose)
int
swiotlb_late_init_with_default_size(size_t default_size)
{
- unsigned long bytes, req_nslabs = io_tlb_nslabs;
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
+ unsigned long bytes, req_nslabs = mem->nslabs;
unsigned char *vstart = NULL;
unsigned int order;
int rc = 0;
- if (!io_tlb_nslabs) {
- io_tlb_nslabs = (default_size >> IO_TLB_SHIFT);
- io_tlb_nslabs = ALIGN(io_tlb_nslabs, IO_TLB_SEGSIZE);
+ if (!mem->nslabs) {
+ mem->nslabs = (default_size >> IO_TLB_SHIFT);
+ mem->nslabs = ALIGN(mem->nslabs, IO_TLB_SEGSIZE);
}
/*
* Get IO TLB memory from the low pages
*/
- order = get_order(io_tlb_nslabs << IO_TLB_SHIFT);
- io_tlb_nslabs = SLABS_PER_PAGE << order;
- bytes = io_tlb_nslabs << IO_TLB_SHIFT;
+ order = get_order(mem->nslabs << IO_TLB_SHIFT);
+ mem->nslabs = SLABS_PER_PAGE << order;
+ bytes = mem->nslabs << IO_TLB_SHIFT;
while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) {
vstart = (void *)__get_free_pages(GFP_DMA | __GFP_NOWARN,
@@ -343,15 +306,15 @@ swiotlb_late_init_with_default_size(size_t default_size)
}
if (!vstart) {
- io_tlb_nslabs = req_nslabs;
+ mem->nslabs = req_nslabs;
return -ENOMEM;
}
if (order != get_order(bytes)) {
pr_warn("only able to allocate %ld MB\n",
(PAGE_SIZE << order) >> 20);
- io_tlb_nslabs = SLABS_PER_PAGE << order;
+ mem->nslabs = SLABS_PER_PAGE << order;
}
- rc = swiotlb_late_init_with_tbl(vstart, io_tlb_nslabs);
+ rc = swiotlb_late_init_with_tbl(vstart, mem->nslabs);
if (rc)
free_pages((unsigned long)vstart, order);
@@ -360,26 +323,32 @@ swiotlb_late_init_with_default_size(size_t default_size)
static void swiotlb_cleanup(void)
{
- io_tlb_end = 0;
- io_tlb_start = 0;
- io_tlb_nslabs = 0;
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
+
+ mem->end = 0;
+ mem->start = 0;
+ mem->nslabs = 0;
max_segment = 0;
}
int
swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
unsigned long i, bytes;
/* protect against double initialization */
- if (WARN_ON_ONCE(io_tlb_start))
+ if (WARN_ON_ONCE(mem->start))
return -ENOMEM;
bytes = nslabs << IO_TLB_SHIFT;
- io_tlb_nslabs = nslabs;
- io_tlb_start = virt_to_phys(tlb);
- io_tlb_end = io_tlb_start + bytes;
+ mem->nslabs = nslabs;
+ mem->start = virt_to_phys(tlb);
+ mem->end = mem->start + bytes;
+ mem->index = 0;
+ mem->late_alloc = 1;
+ spin_lock_init(&mem->lock);
set_memory_decrypted((unsigned long)tlb, bytes >> PAGE_SHIFT);
memset(tlb, 0, bytes);
@@ -387,52 +356,45 @@ swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
/*
* Allocate and initialize the free list array. This array is used
* to find contiguous free memory regions of size up to IO_TLB_SEGSIZE
- * between io_tlb_start and io_tlb_end.
+ * between mem->start and mem->end.
*/
- io_tlb_list = (unsigned int *)__get_free_pages(GFP_KERNEL,
- get_order(io_tlb_nslabs * sizeof(int)));
- if (!io_tlb_list)
+ mem->list = (unsigned int *)__get_free_pages(GFP_KERNEL,
+ get_order(mem->nslabs * sizeof(int)));
+ if (!mem->list)
goto cleanup3;
- io_tlb_orig_addr = (phys_addr_t *)
+ mem->orig_addr = (phys_addr_t *)
__get_free_pages(GFP_KERNEL,
- get_order(io_tlb_nslabs *
+ get_order(mem->nslabs *
sizeof(phys_addr_t)));
- if (!io_tlb_orig_addr)
+ if (!mem->orig_addr)
goto cleanup4;
- io_tlb_alloc_size = (size_t *)
+ mem->alloc_size = (size_t *)
__get_free_pages(GFP_KERNEL,
- get_order(io_tlb_nslabs *
+ get_order(mem->nslabs *
sizeof(size_t)));
- if (!io_tlb_alloc_size)
+ if (!mem->alloc_size)
goto cleanup5;
-
- for (i = 0; i < io_tlb_nslabs; i++) {
- io_tlb_list[i] = IO_TLB_SEGSIZE - io_tlb_offset(i);
- io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
- io_tlb_alloc_size[i] = 0;
+ for (i = 0; i < mem->nslabs; i++) {
+ mem->list[i] = IO_TLB_SEGSIZE - io_tlb_offset(i);
+ mem->orig_addr[i] = INVALID_PHYS_ADDR;
+ mem->alloc_size[i] = 0;
}
- io_tlb_index = 0;
no_iotlb_memory = false;
swiotlb_print_info();
-
- late_alloc = 1;
-
- swiotlb_set_max_segment(io_tlb_nslabs << IO_TLB_SHIFT);
-
+ swiotlb_set_max_segment(mem->nslabs << IO_TLB_SHIFT);
return 0;
cleanup5:
- free_pages((unsigned long)io_tlb_orig_addr, get_order(io_tlb_nslabs *
- sizeof(phys_addr_t)));
-
+ free_pages((unsigned long)mem->orig_addr,
+ get_order(mem->nslabs * sizeof(phys_addr_t)));
cleanup4:
- free_pages((unsigned long)io_tlb_list, get_order(io_tlb_nslabs *
- sizeof(int)));
- io_tlb_list = NULL;
+ free_pages((unsigned long)mem->list,
+ get_order(mem->nslabs * sizeof(int)));
+ mem->list = NULL;
cleanup3:
swiotlb_cleanup();
return -ENOMEM;
@@ -440,27 +402,29 @@ swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
void __init swiotlb_exit(void)
{
- if (!io_tlb_orig_addr)
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
+
+ if (!mem->orig_addr)
return;
- if (late_alloc) {
- free_pages((unsigned long)io_tlb_alloc_size,
- get_order(io_tlb_nslabs * sizeof(size_t)));
- free_pages((unsigned long)io_tlb_orig_addr,
- get_order(io_tlb_nslabs * sizeof(phys_addr_t)));
- free_pages((unsigned long)io_tlb_list, get_order(io_tlb_nslabs *
- sizeof(int)));
- free_pages((unsigned long)phys_to_virt(io_tlb_start),
- get_order(io_tlb_nslabs << IO_TLB_SHIFT));
+ if (mem->late_alloc) {
+ free_pages((unsigned long)mem->alloc_size,
+ get_order(mem->nslabs * sizeof(size_t)));
+ free_pages((unsigned long)mem->orig_addr,
+ get_order(mem->nslabs * sizeof(phys_addr_t)));
+ free_pages((unsigned long)mem->list,
+ get_order(mem->nslabs * sizeof(int)));
+ free_pages((unsigned long)phys_to_virt(mem->start),
+ get_order(mem->nslabs << IO_TLB_SHIFT));
} else {
- memblock_free_late(__pa(io_tlb_orig_addr),
- PAGE_ALIGN(io_tlb_nslabs * sizeof(phys_addr_t)));
- memblock_free_late(__pa(io_tlb_alloc_size),
- PAGE_ALIGN(io_tlb_nslabs * sizeof(size_t)));
- memblock_free_late(__pa(io_tlb_list),
- PAGE_ALIGN(io_tlb_nslabs * sizeof(int)));
- memblock_free_late(io_tlb_start,
- PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT));
+ memblock_free_late(__pa(mem->alloc_size),
+ PAGE_ALIGN(mem->nslabs * sizeof(size_t)));
+ memblock_free_late(__pa(mem->orig_addr),
+ PAGE_ALIGN(mem->nslabs * sizeof(phys_addr_t)));
+ memblock_free_late(__pa(mem->list),
+ PAGE_ALIGN(mem->nslabs * sizeof(int)));
+ memblock_free_late(mem->start,
+ PAGE_ALIGN(mem->nslabs << IO_TLB_SHIFT));
}
swiotlb_cleanup();
}
@@ -471,9 +435,10 @@ void __init swiotlb_exit(void)
static void swiotlb_bounce(struct device *dev, phys_addr_t tlb_addr, size_t size,
enum dma_data_direction dir)
{
- int index = (tlb_addr - io_tlb_start) >> IO_TLB_SHIFT;
- size_t alloc_size = io_tlb_alloc_size[index];
- phys_addr_t orig_addr = io_tlb_orig_addr[index];
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
+ int index = (tlb_addr - mem->start) >> IO_TLB_SHIFT;
+ phys_addr_t orig_addr = mem->orig_addr[index];
+ size_t alloc_size = mem->alloc_size[index];
unsigned long pfn = PFN_DOWN(orig_addr);
unsigned char *vaddr = phys_to_virt(tlb_addr);
@@ -538,9 +503,9 @@ static inline unsigned long get_max_slots(unsigned long boundary_mask)
return nr_slots(boundary_mask + 1);
}
-static unsigned int wrap_index(unsigned int index)
+static unsigned int wrap_index(struct io_tlb_mem *mem, unsigned int index)
{
- if (index >= io_tlb_nslabs)
+ if (index >= mem->nslabs)
return 0;
return index;
}
@@ -552,9 +517,10 @@ static unsigned int wrap_index(unsigned int index)
static int find_slots(struct device *dev, phys_addr_t orig_addr,
size_t alloc_size)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
unsigned long boundary_mask = dma_get_seg_boundary(dev);
dma_addr_t tbl_dma_addr =
- phys_to_dma_unencrypted(dev, io_tlb_start) & boundary_mask;
+ phys_to_dma_unencrypted(dev, mem->start) & boundary_mask;
unsigned long max_slots = get_max_slots(boundary_mask);
unsigned int iotlb_align_mask =
dma_get_min_align_mask(dev) & ~(IO_TLB_SIZE - 1);
@@ -573,15 +539,15 @@ static int find_slots(struct device *dev, phys_addr_t orig_addr,
if (alloc_size >= PAGE_SIZE)
stride = max(stride, stride << (PAGE_SHIFT - IO_TLB_SHIFT));
- spin_lock_irqsave(&io_tlb_lock, flags);
- if (unlikely(nslots > io_tlb_nslabs - io_tlb_used))
+ spin_lock_irqsave(&mem->lock, flags);
+ if (unlikely(nslots > mem->nslabs - mem->used))
goto not_found;
- index = wrap = wrap_index(ALIGN(io_tlb_index, stride));
+ index = wrap = wrap_index(mem, ALIGN(mem->index, stride));
do {
if ((slot_addr(tbl_dma_addr, index) & iotlb_align_mask) !=
(orig_addr & iotlb_align_mask)) {
- index = wrap_index(index + 1);
+ index = wrap_index(mem, index + 1);
continue;
}
@@ -593,34 +559,34 @@ static int find_slots(struct device *dev, phys_addr_t orig_addr,
if (!iommu_is_span_boundary(index, nslots,
nr_slots(tbl_dma_addr),
max_slots)) {
- if (io_tlb_list[index] >= nslots)
+ if (mem->list[index] >= nslots)
goto found;
}
- index = wrap_index(index + stride);
+ index = wrap_index(mem, index + stride);
} while (index != wrap);
not_found:
- spin_unlock_irqrestore(&io_tlb_lock, flags);
+ spin_unlock_irqrestore(&mem->lock, flags);
return -1;
found:
for (i = index; i < index + nslots; i++)
- io_tlb_list[i] = 0;
+ mem->list[i] = 0;
for (i = index - 1;
io_tlb_offset(i) != IO_TLB_SEGSIZE - 1 &&
- io_tlb_list[i]; i--)
- io_tlb_list[i] = ++count;
+ mem->list[i]; i--)
+ mem->list[i] = ++count;
/*
* Update the indices to avoid searching in the next round.
*/
- if (index + nslots < io_tlb_nslabs)
- io_tlb_index = index + nslots;
+ if (index + nslots < mem->nslabs)
+ mem->index = index + nslots;
else
- io_tlb_index = 0;
- io_tlb_used += nslots;
+ mem->index = 0;
+ mem->used += nslots;
- spin_unlock_irqrestore(&io_tlb_lock, flags);
+ spin_unlock_irqrestore(&mem->lock, flags);
return index;
}
@@ -628,6 +594,7 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr,
size_t mapping_size, size_t alloc_size,
enum dma_data_direction dir, unsigned long attrs)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
unsigned int offset = swiotlb_align_offset(dev, orig_addr);
unsigned int index, i;
phys_addr_t tlb_addr;
@@ -649,7 +616,7 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr,
if (!(attrs & DMA_ATTR_NO_WARN))
dev_warn_ratelimited(dev,
"swiotlb buffer is full (sz: %zd bytes), total %lu (slots), used %lu (slots)\n",
- alloc_size, io_tlb_nslabs, io_tlb_used);
+ alloc_size, mem->nslabs, mem->used);
return (phys_addr_t)DMA_MAPPING_ERROR;
}
@@ -659,10 +626,10 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr,
* needed.
*/
for (i = 0; i < nr_slots(alloc_size + offset); i++) {
- io_tlb_orig_addr[index + i] = slot_addr(orig_addr, i);
- io_tlb_alloc_size[index+i] = alloc_size - (i << IO_TLB_SHIFT);
+ mem->orig_addr[index + i] = slot_addr(orig_addr, i);
+ mem->alloc_size[index + i] = alloc_size - (i << IO_TLB_SHIFT);
}
- tlb_addr = slot_addr(io_tlb_start, index) + offset;
+ tlb_addr = slot_addr(mem->start, index) + offset;
if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC) &&
(dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL))
swiotlb_bounce(dev, tlb_addr, mapping_size, DMA_TO_DEVICE);
@@ -676,10 +643,11 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
size_t mapping_size, enum dma_data_direction dir,
unsigned long attrs)
{
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
unsigned long flags;
unsigned int offset = swiotlb_align_offset(hwdev, tlb_addr);
- int index = (tlb_addr - offset - io_tlb_start) >> IO_TLB_SHIFT;
- int nslots = nr_slots(io_tlb_alloc_size[index] + offset);
+ int index = (tlb_addr - offset - mem->start) >> IO_TLB_SHIFT;
+ int nslots = nr_slots(mem->alloc_size[index] + offset);
int count, i;
/*
@@ -695,9 +663,9 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
* While returning the entries to the free list, we merge the entries
* with slots below and above the pool being returned.
*/
- spin_lock_irqsave(&io_tlb_lock, flags);
+ spin_lock_irqsave(&mem->lock, flags);
if (index + nslots < ALIGN(index + 1, IO_TLB_SEGSIZE))
- count = io_tlb_list[index + nslots];
+ count = mem->list[index + nslots];
else
count = 0;
@@ -706,9 +674,9 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
* superceeding slots
*/
for (i = index + nslots - 1; i >= index; i--) {
- io_tlb_list[i] = ++count;
- io_tlb_orig_addr[i] = INVALID_PHYS_ADDR;
- io_tlb_alloc_size[i] = 0;
+ mem->list[i] = ++count;
+ mem->orig_addr[i] = INVALID_PHYS_ADDR;
+ mem->alloc_size[i] = 0;
}
/*
@@ -716,11 +684,11 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
* available (non zero)
*/
for (i = index - 1;
- io_tlb_offset(i) != IO_TLB_SEGSIZE - 1 && io_tlb_list[i];
+ io_tlb_offset(i) != IO_TLB_SEGSIZE - 1 && mem->list[i];
i--)
- io_tlb_list[i] = ++count;
- io_tlb_used -= nslots;
- spin_unlock_irqrestore(&io_tlb_lock, flags);
+ mem->list[i] = ++count;
+ mem->used -= nslots;
+ spin_unlock_irqrestore(&mem->lock, flags);
}
void swiotlb_sync_single_for_device(struct device *dev, phys_addr_t tlb_addr,
@@ -783,21 +751,21 @@ size_t swiotlb_max_mapping_size(struct device *dev)
bool is_swiotlb_active(void)
{
/*
- * When SWIOTLB is initialized, even if io_tlb_start points to physical
- * address zero, io_tlb_end surely doesn't.
+ * When SWIOTLB is initialized, even if mem->start points to physical
+ * address zero, mem->end surely doesn't.
*/
- return io_tlb_end != 0;
+ return io_tlb_default_mem.end != 0;
}
#ifdef CONFIG_DEBUG_FS
static int __init swiotlb_create_debugfs(void)
{
- struct dentry *root;
+ struct io_tlb_mem *mem = &io_tlb_default_mem;
- root = debugfs_create_dir("swiotlb", NULL);
- debugfs_create_ulong("io_tlb_nslabs", 0400, root, &io_tlb_nslabs);
- debugfs_create_ulong("io_tlb_used", 0400, root, &io_tlb_used);
+ mem->debugfs = debugfs_create_dir("swiotlb", NULL);
+ debugfs_create_ulong("io_tlb_nslabs", 0400, mem->debugfs, &mem->nslabs);
+ debugfs_create_ulong("io_tlb_used", 0400, mem->debugfs, &mem->used);
return 0;
}
--
2.29.2
^ permalink raw reply related
* [PATCH 13/14] swiotlb: dynamically allocate io_tlb_default_mem
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
Instead of allocating ->list and ->orig_addr separately just do one
dynamic allocation for the actual io_tlb_mem structure. This simplifies
a lot of the initialization code, and also allows to just check
io_tlb_default_mem to see if swiotlb is in use.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/xen/swiotlb-xen.c | 22 +--
include/linux/swiotlb.h | 18 ++-
kernel/dma/swiotlb.c | 300 +++++++++++++-------------------------
3 files changed, 117 insertions(+), 223 deletions(-)
diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c
index 5329ad54a5f34e..4c89afc0df6289 100644
--- a/drivers/xen/swiotlb-xen.c
+++ b/drivers/xen/swiotlb-xen.c
@@ -158,17 +158,14 @@ static const char *xen_swiotlb_error(enum xen_swiotlb_err err)
int __ref xen_swiotlb_init(void)
{
enum xen_swiotlb_err m_ret = XEN_SWIOTLB_UNKNOWN;
- unsigned long nslabs, bytes, order;
- unsigned int repeat = 3;
+ unsigned long bytes = swiotlb_size_or_default();
+ unsigned long nslabs = bytes >> IO_TLB_SHIFT;
+ unsigned int order, repeat = 3;
int rc = -ENOMEM;
char *start;
- nslabs = swiotlb_nr_tbl();
- if (!nslabs)
- nslabs = DEFAULT_NSLABS;
retry:
m_ret = XEN_SWIOTLB_ENOMEM;
- bytes = nslabs << IO_TLB_SHIFT;
order = get_order(bytes);
/*
@@ -221,19 +218,16 @@ int __ref xen_swiotlb_init(void)
#ifdef CONFIG_X86
void __init xen_swiotlb_init_early(void)
{
- unsigned long nslabs, bytes;
+ unsigned long bytes = swiotlb_size_or_default();
+ unsigned long nslabs = bytes >> IO_TLB_SHIFT;
unsigned int repeat = 3;
char *start;
int rc;
- nslabs = swiotlb_nr_tbl();
- if (!nslabs)
- nslabs = DEFAULT_NSLABS;
retry:
/*
* Get IO TLB memory from any location.
*/
- bytes = nslabs << IO_TLB_SHIFT;
start = memblock_alloc(PAGE_ALIGN(bytes), PAGE_SIZE);
if (!start)
panic("%s: Failed to allocate %lu bytes align=0x%lx\n",
@@ -248,8 +242,8 @@ void __init xen_swiotlb_init_early(void)
if (repeat--) {
/* Min is 2MB */
nslabs = max(1024UL, (nslabs >> 1));
- pr_info("Lowering to %luMB\n",
- (nslabs << IO_TLB_SHIFT) >> 20);
+ bytes = nslabs << IO_TLB_SHIFT;
+ pr_info("Lowering to %luMB\n", bytes >> 20);
goto retry;
}
panic("%s (rc:%d)", xen_swiotlb_error(XEN_SWIOTLB_EFIXUP), rc);
@@ -548,7 +542,7 @@ xen_swiotlb_sync_sg_for_device(struct device *dev, struct scatterlist *sgl,
static int
xen_swiotlb_dma_supported(struct device *hwdev, u64 mask)
{
- return xen_phys_to_dma(hwdev, io_tlb_default_mem.end - 1) <= mask;
+ return xen_phys_to_dma(hwdev, io_tlb_default_mem->end - 1) <= mask;
}
const struct dma_map_ops xen_swiotlb_dma_ops = {
diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
index 5ec5378b17c333..63f7a63f61d098 100644
--- a/include/linux/swiotlb.h
+++ b/include/linux/swiotlb.h
@@ -90,28 +90,30 @@ struct io_tlb_mem {
phys_addr_t end;
unsigned long nslabs;
unsigned long used;
- unsigned int *list;
unsigned int index;
- phys_addr_t *orig_addr;
- size_t *alloc_size;
spinlock_t lock;
struct dentry *debugfs;
bool late_alloc;
+ struct io_tlb_slot {
+ phys_addr_t orig_addr;
+ size_t alloc_size;
+ unsigned int list;
+ } slots[];
};
-extern struct io_tlb_mem io_tlb_default_mem;
+extern struct io_tlb_mem *io_tlb_default_mem;
static inline bool is_swiotlb_buffer(phys_addr_t paddr)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
- return paddr >= mem->start && paddr < mem->end;
+ return mem && paddr >= mem->start && paddr < mem->end;
}
void __init swiotlb_exit(void);
unsigned int swiotlb_max_segment(void);
size_t swiotlb_max_mapping_size(struct device *dev);
bool is_swiotlb_active(void);
-void __init swiotlb_adjust_size(unsigned long new_size);
+void __init swiotlb_adjust_size(unsigned long size);
#else
#define swiotlb_force SWIOTLB_NO_FORCE
static inline bool is_swiotlb_buffer(phys_addr_t paddr)
@@ -135,7 +137,7 @@ static inline bool is_swiotlb_active(void)
return false;
}
-static inline void swiotlb_adjust_size(unsigned long new_size)
+static inline void swiotlb_adjust_size(unsigned long size)
{
}
#endif /* CONFIG_SWIOTLB */
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index 6aa84fa3b1467e..b7bcd7b804bfe8 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -63,7 +63,7 @@
enum swiotlb_force swiotlb_force;
-struct io_tlb_mem io_tlb_default_mem;
+struct io_tlb_mem *io_tlb_default_mem;
/*
* Max segment that we can provide which (if pages are contingous) will
@@ -71,15 +71,15 @@ struct io_tlb_mem io_tlb_default_mem;
*/
static unsigned int max_segment;
+static unsigned long default_nslabs = IO_TLB_DEFAULT_SIZE >> IO_TLB_SHIFT;
+
static int __init
setup_io_tlb_npages(char *str)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
-
if (isdigit(*str)) {
- mem->nslabs = simple_strtoul(str, &str, 0);
/* avoid tail segment of size < IO_TLB_SEGSIZE */
- mem->nslabs = ALIGN(mem->nslabs, IO_TLB_SEGSIZE);
+ default_nslabs =
+ ALIGN(simple_strtoul(str, &str, 0), IO_TLB_SEGSIZE);
}
if (*str == ',')
++str;
@@ -87,24 +87,22 @@ setup_io_tlb_npages(char *str)
swiotlb_force = SWIOTLB_FORCE;
} else if (!strcmp(str, "noforce")) {
swiotlb_force = SWIOTLB_NO_FORCE;
- mem->nslabs = 1;
+ default_nslabs = 1;
}
return 0;
}
early_param("swiotlb", setup_io_tlb_npages);
-static bool no_iotlb_memory;
-
unsigned long swiotlb_nr_tbl(void)
{
- return unlikely(no_iotlb_memory) ? 0 : io_tlb_default_mem.nslabs;
+ return io_tlb_default_mem ? io_tlb_default_mem->nslabs : 0;
}
EXPORT_SYMBOL_GPL(swiotlb_nr_tbl);
unsigned int swiotlb_max_segment(void)
{
- return unlikely(no_iotlb_memory) ? 0 : max_segment;
+ return io_tlb_default_mem ? max_segment : 0;
}
EXPORT_SYMBOL_GPL(swiotlb_max_segment);
@@ -118,44 +116,32 @@ void swiotlb_set_max_segment(unsigned int val)
unsigned long swiotlb_size_or_default(void)
{
- unsigned long size;
-
- size = io_tlb_default_mem.nslabs << IO_TLB_SHIFT;
-
- return size ? size : (IO_TLB_DEFAULT_SIZE);
+ return default_nslabs << IO_TLB_SHIFT;
}
-void __init swiotlb_adjust_size(unsigned long new_size)
+void __init swiotlb_adjust_size(unsigned long size)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
- 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 (!mem->nslabs) {
- size = ALIGN(new_size, IO_TLB_SIZE);
- mem->nslabs = size >> IO_TLB_SHIFT;
- mem->nslabs = ALIGN(mem->nslabs, IO_TLB_SEGSIZE);
-
- pr_info("SWIOTLB bounce buffer size adjusted to %luMB", size >> 20);
- }
+ size = ALIGN(size, IO_TLB_SIZE);
+ default_nslabs = ALIGN(size >> IO_TLB_SHIFT, IO_TLB_SEGSIZE);
+ pr_info("SWIOTLB bounce buffer size adjusted to %luMB", size >> 20);
}
void swiotlb_print_info(void)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
- unsigned long bytes = mem->nslabs << IO_TLB_SHIFT;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
- if (no_iotlb_memory) {
+ if (!mem) {
pr_warn("No low mem\n");
return;
}
pr_info("mapped [mem %pa-%pa] (%luMB)\n", &mem->start, &mem->end,
- bytes >> 20);
+ (mem->nslabs << IO_TLB_SHIFT) >> 20);
}
static inline unsigned long io_tlb_offset(unsigned long val)
@@ -176,13 +162,12 @@ static inline unsigned long nr_slots(u64 val)
*/
void __init swiotlb_update_mem_attributes(void)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
void *vaddr;
unsigned long bytes;
- if (no_iotlb_memory || mem->late_alloc)
+ if (!mem || mem->late_alloc)
return;
-
vaddr = phys_to_virt(mem->start);
bytes = PAGE_ALIGN(mem->nslabs << IO_TLB_SHIFT);
set_memory_decrypted((unsigned long)vaddr, bytes >> PAGE_SHIFT);
@@ -191,49 +176,33 @@ void __init swiotlb_update_mem_attributes(void)
int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
- unsigned long i, bytes;
+ unsigned long bytes = nslabs << IO_TLB_SHIFT, i;
+ struct io_tlb_mem *mem;
size_t alloc_size;
/* protect against double initialization */
- if (WARN_ON_ONCE(mem->start))
+ if (WARN_ON_ONCE(io_tlb_default_mem))
return -ENOMEM;
- bytes = nslabs << IO_TLB_SHIFT;
-
+ alloc_size = PAGE_ALIGN(struct_size(mem, slots, nslabs));
+ mem = memblock_alloc(alloc_size, PAGE_SIZE);
+ if (!mem)
+ panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
+ __func__, alloc_size, PAGE_SIZE);
mem->nslabs = nslabs;
mem->start = __pa(tlb);
mem->end = mem->start + bytes;
mem->index = 0;
spin_lock_init(&mem->lock);
-
- /*
- * Allocate and initialize the free list array. This array is used
- * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE
- * between mem->start and mem->end.
- */
- alloc_size = PAGE_ALIGN(mem->nslabs * sizeof(int));
- mem->list = memblock_alloc(alloc_size, PAGE_SIZE);
- if (!mem->list)
- panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
- __func__, alloc_size, PAGE_SIZE);
-
- alloc_size = PAGE_ALIGN(mem->nslabs * sizeof(phys_addr_t));
- mem->orig_addr = memblock_alloc(alloc_size, PAGE_SIZE);
- if (!mem->orig_addr)
- panic("%s: Failed to allocate %zu bytes align=0x%lx\n",
- __func__, alloc_size, PAGE_SIZE);
-
for (i = 0; i < mem->nslabs; i++) {
- mem->list[i] = IO_TLB_SEGSIZE - io_tlb_offset(i);
- mem->orig_addr[i] = INVALID_PHYS_ADDR;
- mem->alloc_size[i] = 0;
+ mem->slots[i].list = IO_TLB_SEGSIZE - io_tlb_offset(i);
+ mem->slots[i].orig_addr = INVALID_PHYS_ADDR;
+ mem->slots[i].alloc_size = 0;
}
- no_iotlb_memory = false;
+ io_tlb_default_mem = mem;
if (verbose)
swiotlb_print_info();
-
swiotlb_set_max_segment(mem->nslabs << IO_TLB_SHIFT);
return 0;
}
@@ -245,30 +214,21 @@ int __init swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose)
void __init
swiotlb_init(int verbose)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
- size_t default_size = IO_TLB_DEFAULT_SIZE;
- unsigned char *vstart;
- unsigned long bytes;
-
- if (!mem->nslabs) {
- mem->nslabs = (default_size >> IO_TLB_SHIFT);
- mem->nslabs = ALIGN(mem->nslabs, IO_TLB_SEGSIZE);
- }
-
- bytes = mem->nslabs << IO_TLB_SHIFT;
+ size_t bytes = PAGE_ALIGN(default_nslabs << IO_TLB_SHIFT);
+ void *tlb;
/* Get IO TLB memory from the low pages */
- vstart = memblock_alloc_low(PAGE_ALIGN(bytes), PAGE_SIZE);
- if (vstart && !swiotlb_init_with_tbl(vstart, mem->nslabs, verbose))
- return;
-
- if (mem->start) {
- memblock_free_early(mem->start,
- PAGE_ALIGN(mem->nslabs << IO_TLB_SHIFT));
- mem->start = 0;
- }
+ tlb = memblock_alloc_low(bytes, PAGE_SIZE);
+ if (!tlb)
+ goto fail;
+ if (swiotlb_init_with_tbl(tlb, default_nslabs, verbose))
+ goto fail_free_mem;
+ return;
+
+fail_free_mem:
+ memblock_free_early(__pa(tlb), bytes);
+fail:
pr_warn("Cannot allocate buffer");
- no_iotlb_memory = true;
}
/*
@@ -279,23 +239,19 @@ swiotlb_init(int verbose)
int
swiotlb_late_init_with_default_size(size_t default_size)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
- unsigned long bytes, req_nslabs = mem->nslabs;
+ unsigned long nslabs =
+ ALIGN(default_size >> IO_TLB_SHIFT, IO_TLB_SEGSIZE);
+ unsigned long bytes;
unsigned char *vstart = NULL;
unsigned int order;
int rc = 0;
- if (!mem->nslabs) {
- mem->nslabs = (default_size >> IO_TLB_SHIFT);
- mem->nslabs = ALIGN(mem->nslabs, IO_TLB_SEGSIZE);
- }
-
/*
* Get IO TLB memory from the low pages
*/
- order = get_order(mem->nslabs << IO_TLB_SHIFT);
- mem->nslabs = SLABS_PER_PAGE << order;
- bytes = mem->nslabs << IO_TLB_SHIFT;
+ order = get_order(nslabs << IO_TLB_SHIFT);
+ nslabs = SLABS_PER_PAGE << order;
+ bytes = nslabs << IO_TLB_SHIFT;
while ((SLABS_PER_PAGE << order) > IO_TLB_MIN_SLABS) {
vstart = (void *)__get_free_pages(GFP_DMA | __GFP_NOWARN,
@@ -305,43 +261,35 @@ swiotlb_late_init_with_default_size(size_t default_size)
order--;
}
- if (!vstart) {
- mem->nslabs = req_nslabs;
+ if (!vstart)
return -ENOMEM;
- }
+
if (order != get_order(bytes)) {
pr_warn("only able to allocate %ld MB\n",
(PAGE_SIZE << order) >> 20);
- mem->nslabs = SLABS_PER_PAGE << order;
+ nslabs = SLABS_PER_PAGE << order;
}
- rc = swiotlb_late_init_with_tbl(vstart, mem->nslabs);
+ rc = swiotlb_late_init_with_tbl(vstart, nslabs);
if (rc)
free_pages((unsigned long)vstart, order);
return rc;
}
-static void swiotlb_cleanup(void)
-{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
-
- mem->end = 0;
- mem->start = 0;
- mem->nslabs = 0;
- max_segment = 0;
-}
-
int
swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
- unsigned long i, bytes;
+ unsigned long bytes = nslabs << IO_TLB_SHIFT, i;
+ struct io_tlb_mem *mem;
/* protect against double initialization */
- if (WARN_ON_ONCE(mem->start))
+ if (WARN_ON_ONCE(io_tlb_default_mem))
return -ENOMEM;
- bytes = nslabs << IO_TLB_SHIFT;
+ mem = (void *)__get_free_pages(GFP_KERNEL,
+ get_order(struct_size(mem, slots, nslabs)));
+ if (!mem)
+ return -ENOMEM;
mem->nslabs = nslabs;
mem->start = virt_to_phys(tlb);
@@ -349,84 +297,35 @@ swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs)
mem->index = 0;
mem->late_alloc = 1;
spin_lock_init(&mem->lock);
+ for (i = 0; i < mem->nslabs; i++) {
+ mem->slots[i].list = IO_TLB_SEGSIZE - io_tlb_offset(i);
+ mem->slots[i].orig_addr = INVALID_PHYS_ADDR;
+ mem->slots[i].alloc_size = 0;
+ }
set_memory_decrypted((unsigned long)tlb, bytes >> PAGE_SHIFT);
memset(tlb, 0, bytes);
- /*
- * Allocate and initialize the free list array. This array is used
- * to find contiguous free memory regions of size up to IO_TLB_SEGSIZE
- * between mem->start and mem->end.
- */
- mem->list = (unsigned int *)__get_free_pages(GFP_KERNEL,
- get_order(mem->nslabs * sizeof(int)));
- if (!mem->list)
- goto cleanup3;
-
- mem->orig_addr = (phys_addr_t *)
- __get_free_pages(GFP_KERNEL,
- get_order(mem->nslabs *
- sizeof(phys_addr_t)));
- if (!mem->orig_addr)
- goto cleanup4;
-
- mem->alloc_size = (size_t *)
- __get_free_pages(GFP_KERNEL,
- get_order(mem->nslabs *
- sizeof(size_t)));
- if (!mem->alloc_size)
- goto cleanup5;
-
- for (i = 0; i < mem->nslabs; i++) {
- mem->list[i] = IO_TLB_SEGSIZE - io_tlb_offset(i);
- mem->orig_addr[i] = INVALID_PHYS_ADDR;
- mem->alloc_size[i] = 0;
- }
- no_iotlb_memory = false;
-
+ io_tlb_default_mem = mem;
swiotlb_print_info();
swiotlb_set_max_segment(mem->nslabs << IO_TLB_SHIFT);
return 0;
-
-cleanup5:
- free_pages((unsigned long)mem->orig_addr,
- get_order(mem->nslabs * sizeof(phys_addr_t)));
-cleanup4:
- free_pages((unsigned long)mem->list,
- get_order(mem->nslabs * sizeof(int)));
- mem->list = NULL;
-cleanup3:
- swiotlb_cleanup();
- return -ENOMEM;
}
void __init swiotlb_exit(void)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
+ size_t size;
- if (!mem->orig_addr)
+ if (!mem)
return;
- if (mem->late_alloc) {
- free_pages((unsigned long)mem->alloc_size,
- get_order(mem->nslabs * sizeof(size_t)));
- free_pages((unsigned long)mem->orig_addr,
- get_order(mem->nslabs * sizeof(phys_addr_t)));
- free_pages((unsigned long)mem->list,
- get_order(mem->nslabs * sizeof(int)));
- free_pages((unsigned long)phys_to_virt(mem->start),
- get_order(mem->nslabs << IO_TLB_SHIFT));
- } else {
- memblock_free_late(__pa(mem->alloc_size),
- PAGE_ALIGN(mem->nslabs * sizeof(size_t)));
- memblock_free_late(__pa(mem->orig_addr),
- PAGE_ALIGN(mem->nslabs * sizeof(phys_addr_t)));
- memblock_free_late(__pa(mem->list),
- PAGE_ALIGN(mem->nslabs * sizeof(int)));
- memblock_free_late(mem->start,
- PAGE_ALIGN(mem->nslabs << IO_TLB_SHIFT));
- }
- swiotlb_cleanup();
+ size = struct_size(mem, slots, mem->nslabs);
+ if (mem->late_alloc)
+ free_pages((unsigned long)mem, get_order(size));
+ else
+ memblock_free_late(__pa(mem), PAGE_ALIGN(size));
+ io_tlb_default_mem = NULL;
}
/*
@@ -435,10 +334,10 @@ void __init swiotlb_exit(void)
static void swiotlb_bounce(struct device *dev, phys_addr_t tlb_addr, size_t size,
enum dma_data_direction dir)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
int index = (tlb_addr - mem->start) >> IO_TLB_SHIFT;
- phys_addr_t orig_addr = mem->orig_addr[index];
- size_t alloc_size = mem->alloc_size[index];
+ phys_addr_t orig_addr = mem->slots[index].orig_addr;
+ size_t alloc_size = mem->slots[index].alloc_size;
unsigned long pfn = PFN_DOWN(orig_addr);
unsigned char *vaddr = phys_to_virt(tlb_addr);
@@ -517,7 +416,7 @@ static unsigned int wrap_index(struct io_tlb_mem *mem, unsigned int index)
static int find_slots(struct device *dev, phys_addr_t orig_addr,
size_t alloc_size)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
unsigned long boundary_mask = dma_get_seg_boundary(dev);
dma_addr_t tbl_dma_addr =
phys_to_dma_unencrypted(dev, mem->start) & boundary_mask;
@@ -559,7 +458,7 @@ static int find_slots(struct device *dev, phys_addr_t orig_addr,
if (!iommu_is_span_boundary(index, nslots,
nr_slots(tbl_dma_addr),
max_slots)) {
- if (mem->list[index] >= nslots)
+ if (mem->slots[index].list >= nslots)
goto found;
}
index = wrap_index(mem, index + stride);
@@ -571,11 +470,11 @@ static int find_slots(struct device *dev, phys_addr_t orig_addr,
found:
for (i = index; i < index + nslots; i++)
- mem->list[i] = 0;
+ mem->slots[i].list = 0;
for (i = index - 1;
io_tlb_offset(i) != IO_TLB_SEGSIZE - 1 &&
- mem->list[i]; i--)
- mem->list[i] = ++count;
+ mem->slots[i].list; i--)
+ mem->slots[i].list = ++count;
/*
* Update the indices to avoid searching in the next round.
@@ -594,12 +493,12 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr,
size_t mapping_size, size_t alloc_size,
enum dma_data_direction dir, unsigned long attrs)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
unsigned int offset = swiotlb_align_offset(dev, orig_addr);
unsigned int index, i;
phys_addr_t tlb_addr;
- if (no_iotlb_memory)
+ if (!mem)
panic("Can not allocate SWIOTLB buffer earlier and can't now provide you with the DMA bounce buffer");
if (mem_encrypt_active())
@@ -626,8 +525,9 @@ phys_addr_t swiotlb_tbl_map_single(struct device *dev, phys_addr_t orig_addr,
* needed.
*/
for (i = 0; i < nr_slots(alloc_size + offset); i++) {
- mem->orig_addr[index + i] = slot_addr(orig_addr, i);
- mem->alloc_size[index + i] = alloc_size - (i << IO_TLB_SHIFT);
+ mem->slots[index + i].orig_addr = slot_addr(orig_addr, i);
+ mem->slots[index + i].alloc_size =
+ alloc_size - (i << IO_TLB_SHIFT);
}
tlb_addr = slot_addr(mem->start, index) + offset;
if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC) &&
@@ -643,11 +543,11 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
size_t mapping_size, enum dma_data_direction dir,
unsigned long attrs)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
unsigned long flags;
unsigned int offset = swiotlb_align_offset(hwdev, tlb_addr);
int index = (tlb_addr - offset - mem->start) >> IO_TLB_SHIFT;
- int nslots = nr_slots(mem->alloc_size[index] + offset);
+ int nslots = nr_slots(mem->slots[index].alloc_size + offset);
int count, i;
/*
@@ -665,7 +565,7 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
*/
spin_lock_irqsave(&mem->lock, flags);
if (index + nslots < ALIGN(index + 1, IO_TLB_SEGSIZE))
- count = mem->list[index + nslots];
+ count = mem->slots[index + nslots].list;
else
count = 0;
@@ -674,9 +574,9 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
* superceeding slots
*/
for (i = index + nslots - 1; i >= index; i--) {
- mem->list[i] = ++count;
- mem->orig_addr[i] = INVALID_PHYS_ADDR;
- mem->alloc_size[i] = 0;
+ mem->slots[i].list = ++count;
+ mem->slots[i].orig_addr = INVALID_PHYS_ADDR;
+ mem->slots[i].alloc_size = 0;
}
/*
@@ -684,9 +584,9 @@ void swiotlb_tbl_unmap_single(struct device *hwdev, phys_addr_t tlb_addr,
* available (non zero)
*/
for (i = index - 1;
- io_tlb_offset(i) != IO_TLB_SEGSIZE - 1 && mem->list[i];
+ io_tlb_offset(i) != IO_TLB_SEGSIZE - 1 && mem->slots[i].list;
i--)
- mem->list[i] = ++count;
+ mem->slots[i].list = ++count;
mem->used -= nslots;
spin_unlock_irqrestore(&mem->lock, flags);
}
@@ -750,19 +650,17 @@ size_t swiotlb_max_mapping_size(struct device *dev)
bool is_swiotlb_active(void)
{
- /*
- * When SWIOTLB is initialized, even if mem->start points to physical
- * address zero, mem->end surely doesn't.
- */
- return io_tlb_default_mem.end != 0;
+ return io_tlb_default_mem != NULL;
}
#ifdef CONFIG_DEBUG_FS
static int __init swiotlb_create_debugfs(void)
{
- struct io_tlb_mem *mem = &io_tlb_default_mem;
+ struct io_tlb_mem *mem = io_tlb_default_mem;
+ if (!mem)
+ return 0;
mem->debugfs = debugfs_create_dir("swiotlb", NULL);
debugfs_create_ulong("io_tlb_nslabs", 0400, mem->debugfs, &mem->nslabs);
debugfs_create_ulong("io_tlb_used", 0400, mem->debugfs, &mem->used);
--
2.29.2
^ permalink raw reply related
* [PATCH 14/14] swiotlb: remove swiotlb_nr_tbl
From: Christoph Hellwig @ 2021-03-01 7:44 UTC (permalink / raw)
To: Konrad Rzeszutek Wilk
Cc: iommu, xen-devel, Claire Chang, linuxppc-dev, Dongli Zhang
In-Reply-To: <20210301074436.919889-1-hch@lst.de>
All callers just use it to check if swiotlb is active at all, for which
they can just use is_swiotlb_active. In the longer run drivers need
to stop using is_swiotlb_active as well, but let's do the simple step
first.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/gpu/drm/i915/gem/i915_gem_internal.c | 2 +-
drivers/gpu/drm/nouveau/nouveau_ttm.c | 2 +-
drivers/pci/xen-pcifront.c | 2 +-
include/linux/swiotlb.h | 1 -
kernel/dma/swiotlb.c | 7 +------
5 files changed, 4 insertions(+), 10 deletions(-)
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_internal.c b/drivers/gpu/drm/i915/gem/i915_gem_internal.c
index ad22f42541bda6..a9d65fc8aa0eab 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_internal.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_internal.c
@@ -42,7 +42,7 @@ static int i915_gem_object_get_pages_internal(struct drm_i915_gem_object *obj)
max_order = MAX_ORDER;
#ifdef CONFIG_SWIOTLB
- if (swiotlb_nr_tbl()) {
+ if (is_swiotlb_active()) {
unsigned int max_segment;
max_segment = swiotlb_max_segment();
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index a37bc3d7b38b3b..9662522aa0664a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -321,7 +321,7 @@ nouveau_ttm_init(struct nouveau_drm *drm)
}
#if IS_ENABLED(CONFIG_SWIOTLB) && IS_ENABLED(CONFIG_X86)
- need_swiotlb = !!swiotlb_nr_tbl();
+ need_swiotlb = is_swiotlb_active();
#endif
ret = ttm_bo_device_init(&drm->ttm.bdev, &nouveau_bo_driver,
diff --git a/drivers/pci/xen-pcifront.c b/drivers/pci/xen-pcifront.c
index c6fe0cfec0f681..a549e822033fd6 100644
--- a/drivers/pci/xen-pcifront.c
+++ b/drivers/pci/xen-pcifront.c
@@ -693,7 +693,7 @@ static int pcifront_connect_and_init_dma(struct pcifront_device *pdev)
spin_unlock(&pcifront_dev_lock);
- if (!err && !swiotlb_nr_tbl()) {
+ if (!err && !is_swiotlb_active()) {
err = pci_xen_swiotlb_init_late();
if (err)
dev_err(&pdev->xdev->dev, "Could not setup SWIOTLB!\n");
diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h
index 63f7a63f61d098..216854a5e5134b 100644
--- a/include/linux/swiotlb.h
+++ b/include/linux/swiotlb.h
@@ -37,7 +37,6 @@ enum swiotlb_force {
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 int swiotlb_late_init_with_tbl(char *tlb, unsigned long nslabs);
extern int swiotlb_late_init_with_default_size(size_t default_size);
diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c
index b7bcd7b804bfe8..809d5fdc204675 100644
--- a/kernel/dma/swiotlb.c
+++ b/kernel/dma/swiotlb.c
@@ -94,12 +94,6 @@ setup_io_tlb_npages(char *str)
}
early_param("swiotlb", setup_io_tlb_npages);
-unsigned long swiotlb_nr_tbl(void)
-{
- return io_tlb_default_mem ? io_tlb_default_mem->nslabs : 0;
-}
-EXPORT_SYMBOL_GPL(swiotlb_nr_tbl);
-
unsigned int swiotlb_max_segment(void)
{
return io_tlb_default_mem ? max_segment : 0;
@@ -652,6 +646,7 @@ bool is_swiotlb_active(void)
{
return io_tlb_default_mem != NULL;
}
+EXPORT_SYMBOL_GPL(is_swiotlb_active);
#ifdef CONFIG_DEBUG_FS
--
2.29.2
^ permalink raw reply related
* Re: [PATCH] mm: Generalize HUGETLB_PAGE_SIZE_VARIABLE
From: Christoph Hellwig @ 2021-03-01 7:53 UTC (permalink / raw)
To: Anshuman Khandual
Cc: linux-ia64, linux-kernel, linux-mm, Paul Mackerras, Andrew Morton,
linuxppc-dev, Christoph Hellwig
In-Reply-To: <89f2d77c-f4bc-8f7b-a6b0-1c04e422fb77@arm.com>
On Mon, Mar 01, 2021 at 01:13:41PM +0530, Anshuman Khandual wrote:
> > doesn't this need a 'if HUGETLB_PAGE'
>
> While making HUGETLB_PAGE_SIZE_VARIABLE a generic option, also made it
> dependent on HUGETLB_PAGE. Should not that gate HUGETLB_PAGE_SIZE_VARIABLE
> when HUGETLB_PAGE is not available irrespective of the select statement on
> the platforms ?
depends doesn't properly work for variables that are selected.
^ permalink raw reply
* Re: [PATCH] mm: Generalize HUGETLB_PAGE_SIZE_VARIABLE
From: Anshuman Khandual @ 2021-03-01 8:01 UTC (permalink / raw)
To: Christoph Hellwig
Cc: linux-ia64, linux-kernel, linux-mm, Paul Mackerras, Andrew Morton,
linuxppc-dev
In-Reply-To: <20210301075352.GA27659@lst.de>
On 3/1/21 1:23 PM, Christoph Hellwig wrote:
> On Mon, Mar 01, 2021 at 01:13:41PM +0530, Anshuman Khandual wrote:
>>> doesn't this need a 'if HUGETLB_PAGE'
>>
>> While making HUGETLB_PAGE_SIZE_VARIABLE a generic option, also made it
>> dependent on HUGETLB_PAGE. Should not that gate HUGETLB_PAGE_SIZE_VARIABLE
>> when HUGETLB_PAGE is not available irrespective of the select statement on
>> the platforms ?
>
> depends doesn't properly work for variables that are selected.
>
Alright, will move the HUGETLB_PAGE dependency to platforms while selecting
the variable HUGETLB_PAGE_SIZE_VARIABLE.
^ permalink raw reply
* cleanup unused or almost unused IOMMU APIs and the FSL PAMU driver
From: Christoph Hellwig @ 2021-03-01 8:42 UTC (permalink / raw)
To: Joerg Roedel, Will Deacon, Li Yang
Cc: freedreno, kvm, linuxppc-dev, dri-devel, virtualization, iommu,
netdev, linux-arm-msm, David Woodhouse, linux-arm-kernel,
Lu Baolu
Hi all,
there are a bunch of IOMMU APIs that are entirely unused, or only used as
a private communication channel between the FSL PAMU driver and it's only
consumer, the qbman portal driver.
So this series drops a huge chunk of entirely unused FSL PAMU
functionality, then drops all kinds of unused IOMMU APIs, and then
replaces what is left of the iommu_attrs with properly typed, smaller
and easier to use specific APIs.
Diffstat:
arch/powerpc/include/asm/fsl_pamu_stash.h | 12
drivers/gpu/drm/msm/adreno/adreno_gpu.c | 2
drivers/iommu/amd/iommu.c | 23
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 85 ---
drivers/iommu/arm/arm-smmu/arm-smmu.c | 122 +---
drivers/iommu/dma-iommu.c | 8
drivers/iommu/fsl_pamu.c | 264 ----------
drivers/iommu/fsl_pamu.h | 10
drivers/iommu/fsl_pamu_domain.c | 694 ++--------------------------
drivers/iommu/fsl_pamu_domain.h | 46 -
drivers/iommu/intel/iommu.c | 55 --
drivers/iommu/iommu.c | 75 ---
drivers/soc/fsl/qbman/qman_portal.c | 56 --
drivers/vfio/vfio_iommu_type1.c | 31 -
drivers/vhost/vdpa.c | 10
include/linux/iommu.h | 81 ---
16 files changed, 214 insertions(+), 1360 deletions(-)
^ permalink raw reply
* [PATCH 01/17] iommu: remove the unused domain_window_disable method
From: Christoph Hellwig @ 2021-03-01 8:42 UTC (permalink / raw)
To: Joerg Roedel, Will Deacon, Li Yang
Cc: freedreno, kvm, linuxppc-dev, dri-devel, virtualization, iommu,
netdev, linux-arm-msm, David Woodhouse, linux-arm-kernel,
Lu Baolu
In-Reply-To: <20210301084257.945454-1-hch@lst.de>
domain_window_disable is wired up by fsl_pamu, but never actually called.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/iommu/fsl_pamu_domain.c | 48 ---------------------------------
include/linux/iommu.h | 2 --
2 files changed, 50 deletions(-)
diff --git a/drivers/iommu/fsl_pamu_domain.c b/drivers/iommu/fsl_pamu_domain.c
index b2110767caf49c..53380cf1fa452f 100644
--- a/drivers/iommu/fsl_pamu_domain.c
+++ b/drivers/iommu/fsl_pamu_domain.c
@@ -473,53 +473,6 @@ static int update_domain_mapping(struct fsl_dma_domain *dma_domain, u32 wnd_nr)
return ret;
}
-static int disable_domain_win(struct fsl_dma_domain *dma_domain, u32 wnd_nr)
-{
- struct device_domain_info *info;
- int ret = 0;
-
- list_for_each_entry(info, &dma_domain->devices, link) {
- if (dma_domain->win_cnt == 1 && dma_domain->enabled) {
- ret = pamu_disable_liodn(info->liodn);
- if (!ret)
- dma_domain->enabled = 0;
- } else {
- ret = pamu_disable_spaace(info->liodn, wnd_nr);
- }
- }
-
- return ret;
-}
-
-static void fsl_pamu_window_disable(struct iommu_domain *domain, u32 wnd_nr)
-{
- struct fsl_dma_domain *dma_domain = to_fsl_dma_domain(domain);
- unsigned long flags;
- int ret;
-
- spin_lock_irqsave(&dma_domain->domain_lock, flags);
- if (!dma_domain->win_arr) {
- pr_debug("Number of windows not configured\n");
- spin_unlock_irqrestore(&dma_domain->domain_lock, flags);
- return;
- }
-
- if (wnd_nr >= dma_domain->win_cnt) {
- pr_debug("Invalid window index\n");
- spin_unlock_irqrestore(&dma_domain->domain_lock, flags);
- return;
- }
-
- if (dma_domain->win_arr[wnd_nr].valid) {
- ret = disable_domain_win(dma_domain, wnd_nr);
- if (!ret) {
- dma_domain->win_arr[wnd_nr].valid = 0;
- dma_domain->mapped--;
- }
- }
-
- spin_unlock_irqrestore(&dma_domain->domain_lock, flags);
-}
static int fsl_pamu_window_enable(struct iommu_domain *domain, u32 wnd_nr,
phys_addr_t paddr, u64 size, int prot)
@@ -1032,7 +985,6 @@ static const struct iommu_ops fsl_pamu_ops = {
.attach_dev = fsl_pamu_attach_device,
.detach_dev = fsl_pamu_detach_device,
.domain_window_enable = fsl_pamu_window_enable,
- .domain_window_disable = fsl_pamu_window_disable,
.iova_to_phys = fsl_pamu_iova_to_phys,
.domain_set_attr = fsl_pamu_set_domain_attr,
.domain_get_attr = fsl_pamu_get_domain_attr,
diff --git a/include/linux/iommu.h b/include/linux/iommu.h
index 5e7fe519430af4..47c8b318d8f523 100644
--- a/include/linux/iommu.h
+++ b/include/linux/iommu.h
@@ -209,7 +209,6 @@ struct iommu_iotlb_gather {
* @put_resv_regions: Free list of reserved regions for a device
* @apply_resv_region: Temporary helper call-back for iova reserved ranges
* @domain_window_enable: Configure and enable a particular window for a domain
- * @domain_window_disable: Disable a particular window for a domain
* @of_xlate: add OF master IDs to iommu grouping
* @is_attach_deferred: Check if domain attach should be deferred from iommu
* driver init to device driver init (default no)
@@ -270,7 +269,6 @@ struct iommu_ops {
/* Window handling functions */
int (*domain_window_enable)(struct iommu_domain *domain, u32 wnd_nr,
phys_addr_t paddr, u64 size, int prot);
- void (*domain_window_disable)(struct iommu_domain *domain, u32 wnd_nr);
int (*of_xlate)(struct device *dev, struct of_phandle_args *args);
bool (*is_attach_deferred)(struct iommu_domain *domain, struct device *dev);
--
2.29.2
^ permalink raw reply related
* [PATCH 02/17] iommu/fsl_pamu: remove fsl_pamu_get_domain_attr
From: Christoph Hellwig @ 2021-03-01 8:42 UTC (permalink / raw)
To: Joerg Roedel, Will Deacon, Li Yang
Cc: freedreno, kvm, linuxppc-dev, dri-devel, virtualization, iommu,
netdev, linux-arm-msm, David Woodhouse, linux-arm-kernel,
Lu Baolu
In-Reply-To: <20210301084257.945454-1-hch@lst.de>
None of the values returned by this function are ever queried. Also
remove the DOMAIN_ATTR_FSL_PAMUV1 enum value that is not otherwise used.
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
drivers/iommu/fsl_pamu_domain.c | 30 ------------------------------
include/linux/iommu.h | 4 ----
2 files changed, 34 deletions(-)
diff --git a/drivers/iommu/fsl_pamu_domain.c b/drivers/iommu/fsl_pamu_domain.c
index 53380cf1fa452f..e587ec43f7e750 100644
--- a/drivers/iommu/fsl_pamu_domain.c
+++ b/drivers/iommu/fsl_pamu_domain.c
@@ -832,35 +832,6 @@ static int fsl_pamu_set_domain_attr(struct iommu_domain *domain,
return ret;
}
-static int fsl_pamu_get_domain_attr(struct iommu_domain *domain,
- enum iommu_attr attr_type, void *data)
-{
- struct fsl_dma_domain *dma_domain = to_fsl_dma_domain(domain);
- int ret = 0;
-
- switch (attr_type) {
- case DOMAIN_ATTR_FSL_PAMU_STASH:
- memcpy(data, &dma_domain->dma_stash,
- sizeof(struct pamu_stash_attribute));
- break;
- case DOMAIN_ATTR_FSL_PAMU_ENABLE:
- *(int *)data = dma_domain->enabled;
- break;
- case DOMAIN_ATTR_FSL_PAMUV1:
- *(int *)data = DOMAIN_ATTR_FSL_PAMUV1;
- break;
- case DOMAIN_ATTR_WINDOWS:
- *(u32 *)data = dma_domain->win_cnt;
- break;
- default:
- pr_debug("Unsupported attribute type\n");
- ret = -EINVAL;
- break;
- }
-
- return ret;
-}
-
static struct iommu_group *get_device_iommu_group(struct device *dev)
{
struct iommu_group *group;
@@ -987,7 +958,6 @@ static const struct iommu_ops fsl_pamu_ops = {
.domain_window_enable = fsl_pamu_window_enable,
.iova_to_phys = fsl_pamu_iova_to_phys,
.domain_set_attr = fsl_pamu_set_domain_attr,
- .domain_get_attr = fsl_pamu_get_domain_attr,
.probe_device = fsl_pamu_probe_device,
.release_device = fsl_pamu_release_device,
.device_group = fsl_pamu_device_group,
diff --git a/include/linux/iommu.h b/include/linux/iommu.h
index 47c8b318d8f523..52874ae164dd60 100644
--- a/include/linux/iommu.h
+++ b/include/linux/iommu.h
@@ -104,9 +104,6 @@ enum iommu_cap {
* -the actual size of the mapped region of a window must be power
* of 2 starting with 4KB and physical address must be naturally
* aligned.
- * DOMAIN_ATTR_FSL_PAMUV1 corresponds to the above mentioned contraints.
- * The caller can invoke iommu_domain_get_attr to check if the underlying
- * iommu implementation supports these constraints.
*/
enum iommu_attr {
@@ -115,7 +112,6 @@ enum iommu_attr {
DOMAIN_ATTR_WINDOWS,
DOMAIN_ATTR_FSL_PAMU_STASH,
DOMAIN_ATTR_FSL_PAMU_ENABLE,
- DOMAIN_ATTR_FSL_PAMUV1,
DOMAIN_ATTR_NESTING, /* two stages of translation */
DOMAIN_ATTR_DMA_USE_FLUSH_QUEUE,
DOMAIN_ATTR_IO_PGTABLE_CFG,
--
2.29.2
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox