[RFC,v3,3/4] Introduce cpu_dcache_is_aliasing() across all architectures
Commit Message
Introduce a generic way to query whether the data cache is virtually
aliased on all architectures. Its purpose is to ensure that subsystems
which are incompatible with virtually aliased data caches (e.g. FS_DAX)
can reliably query this.
For data cache aliasing, there are three scenarios dependending on the
architecture. Here is a breakdown based on my understanding:
A) The data cache is always aliasing:
* arc
* csky
* m68k (note: shared memory mappings are incoherent ? SHMLBA is missing there.)
* sh
* parisc
B) The data cache aliasing is statically known or depends on querying CPU
state at runtime:
* arm (cache_is_vivt() || cache_is_vipt_aliasing())
* mips (cpu_has_dc_aliases)
* nios2 (NIOS2_DCACHE_SIZE > PAGE_SIZE)
* sparc32 (vac_cache_size > PAGE_SIZE)
* sparc64 (L1DCACHE_SIZE > PAGE_SIZE)
* xtensa (DCACHE_WAY_SIZE > PAGE_SIZE)
C) The data cache is never aliasing:
* alpha
* arm64 (aarch64)
* hexagon
* loongarch (but with incoherent write buffers, which are disabled since
commit d23b7795 ("LoongArch: Change SHMLBA from SZ_64K to PAGE_SIZE"))
* microblaze
* openrisc
* powerpc
* riscv
* s390
* um
* x86
Require architectures in A) and B) to select ARCH_HAS_CPU_CACHE_ALIASING and
implement "cpu_dcache_is_aliasing()".
Architectures in C) don't select ARCH_HAS_CPU_CACHE_ALIASING, and thus
cpu_dcache_is_aliasing() simply evaluates to "false".
Note that this leaves "cpu_icache_is_aliasing()" to be implemented as future
work. This would be useful to gate features like XIP on architectures
which have aliasing CPU dcache-icache but not CPU dcache-dcache.
Use "cpu_dcache" and "cpu_cache" rather than just "dcache" and "cache"
to clarify that we really mean "CPU data cache" and "CPU cache" to
eliminate any possible confusion with VFS "dentry cache" and "page
cache".
Link: https://lore.kernel.org/lkml/20030910210416.GA24258@mail.jlokier.co.uk/
Fixes: d92576f1167c ("dax: does not work correctly with virtual aliasing caches")
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-mm@kvack.org
Cc: linux-arch@vger.kernel.org
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Russell King <linux@armlinux.org.uk>
Cc: linux-cxl@vger.kernel.org
Cc: nvdimm@lists.linux.dev
Cc: linux-xfs@vger.kernel.org
Cc: dm-devel@lists.linux.dev
---
arch/arc/Kconfig | 1 +
arch/arc/include/asm/cachetype.h | 9 +++++++++
arch/arm/Kconfig | 1 +
arch/arm/include/asm/cachetype.h | 2 ++
arch/csky/Kconfig | 1 +
arch/csky/include/asm/cachetype.h | 9 +++++++++
arch/m68k/Kconfig | 1 +
arch/m68k/include/asm/cachetype.h | 9 +++++++++
arch/mips/Kconfig | 1 +
arch/mips/include/asm/cachetype.h | 9 +++++++++
arch/nios2/Kconfig | 1 +
arch/nios2/include/asm/cachetype.h | 10 ++++++++++
arch/parisc/Kconfig | 1 +
arch/parisc/include/asm/cachetype.h | 9 +++++++++
arch/sh/Kconfig | 1 +
arch/sh/include/asm/cachetype.h | 9 +++++++++
arch/sparc/Kconfig | 1 +
arch/sparc/include/asm/cachetype.h | 14 ++++++++++++++
arch/xtensa/Kconfig | 1 +
arch/xtensa/include/asm/cachetype.h | 10 ++++++++++
include/linux/cacheinfo.h | 6 ++++++
mm/Kconfig | 6 ++++++
22 files changed, 112 insertions(+)
create mode 100644 arch/arc/include/asm/cachetype.h
create mode 100644 arch/csky/include/asm/cachetype.h
create mode 100644 arch/m68k/include/asm/cachetype.h
create mode 100644 arch/mips/include/asm/cachetype.h
create mode 100644 arch/nios2/include/asm/cachetype.h
create mode 100644 arch/parisc/include/asm/cachetype.h
create mode 100644 arch/sh/include/asm/cachetype.h
create mode 100644 arch/sparc/include/asm/cachetype.h
create mode 100644 arch/xtensa/include/asm/cachetype.h
Comments
So this is the third iteration and you still keep only sending patch
3 to the list. How is anyone supposed to review it if you don't send
them all the pieces?
On 2024-01-31 12:17, Christoph Hellwig wrote:
> So this is the third iteration and you still keep only sending patch
> 3 to the list. How is anyone supposed to review it if you don't send
> them all the pieces?
My bad. I was aiming for not spamming mailing lists on unrelated patches. I did
not CC linux-xfs@vger.kernel.org on the other patches. But the missing
context is just confusing. And I forgot to CC linux-fsdevel@vger.kernel.org
on patch 3 as well.
You can find the entire series on lore:
https://lore.kernel.org/lkml/20240131162533.247710-1-mathieu.desnoyers@efficios.com/T/#t
I'll make sure to copy all lists for all patches in the next
round, namely:
Cc: linux-arch@vger.kernel.org
Cc: linux-cxl@vger.kernel.org
Cc: linux-fsdevel@vger.kernel.org
Cc: linux-mm@kvack.org
Cc: linux-xfs@vger.kernel.org
Cc: dm-devel@lists.linux.dev
Cc: nvdimm@lists.linux.dev
Thanks,
Mathieu
Mathieu Desnoyers wrote:
> On 2024-01-31 12:17, Christoph Hellwig wrote:
> > So this is the third iteration and you still keep only sending patch
> > 3 to the list. How is anyone supposed to review it if you don't send
> > them all the pieces?
>
> My bad. I was aiming for not spamming mailing lists on unrelated patches. I did
> not CC linux-xfs@vger.kernel.org on the other patches. But the missing
> context is just confusing. And I forgot to CC linux-fsdevel@vger.kernel.org
> on patch 3 as well.
>
> You can find the entire series on lore:
>
> https://lore.kernel.org/lkml/20240131162533.247710-1-mathieu.desnoyers@efficios.com/T/#t
>
> I'll make sure to copy all lists for all patches in the next
> round, namely:
>
> Cc: linux-arch@vger.kernel.org
> Cc: linux-cxl@vger.kernel.org
> Cc: linux-fsdevel@vger.kernel.org
> Cc: linux-mm@kvack.org
> Cc: linux-xfs@vger.kernel.org
> Cc: dm-devel@lists.linux.dev
> Cc: nvdimm@lists.linux.dev
>
> Thanks,
>
> Mathieu
FWIW there are some developers and lists that want all patches and there
are some developers only want to see the one patch they are directly
copied on. I have a locally maintained list as I have discovered these
different preferences, but maybe MAINTAINERS could carry a flag to
indicate this to save the next person some time?
@@ -6,6 +6,7 @@
config ARC
def_bool y
select ARC_TIMERS
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_HAS_CACHE_LINE_SIZE
select ARCH_HAS_DEBUG_VM_PGTABLE
select ARCH_HAS_DMA_PREP_COHERENT
new file mode 100644
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_ARC_CACHETYPE_H
+#define __ASM_ARC_CACHETYPE_H
+
+#include <linux/types.h>
+
+#define cpu_dcache_is_aliasing() true
+
+#endif
@@ -5,6 +5,7 @@ config ARM
select ARCH_32BIT_OFF_T
select ARCH_CORRECT_STACKTRACE_ON_KRETPROBE if HAVE_KRETPROBES && FRAME_POINTER && !ARM_UNWIND
select ARCH_HAS_BINFMT_FLAT
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_HAS_CPU_FINALIZE_INIT if MMU
select ARCH_HAS_CURRENT_STACK_POINTER
select ARCH_HAS_DEBUG_VIRTUAL if MMU
@@ -20,6 +20,8 @@ extern unsigned int cacheid;
#define icache_is_vipt_aliasing() cacheid_is(CACHEID_VIPT_I_ALIASING)
#define icache_is_pipt() cacheid_is(CACHEID_PIPT)
+#define cpu_dcache_is_aliasing() (cache_is_vivt() || cache_is_vipt_aliasing())
+
/*
* __LINUX_ARM_ARCH__ is the minimum supported CPU architecture
* Mask out support which will never be present on newer CPUs.
@@ -2,6 +2,7 @@
config CSKY
def_bool y
select ARCH_32BIT_OFF_T
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_HAS_DMA_PREP_COHERENT
select ARCH_HAS_GCOV_PROFILE_ALL
select ARCH_HAS_SYNC_DMA_FOR_CPU
new file mode 100644
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_CSKY_CACHETYPE_H
+#define __ASM_CSKY_CACHETYPE_H
+
+#include <linux/types.h>
+
+#define cpu_dcache_is_aliasing() true
+
+#endif
@@ -3,6 +3,7 @@ config M68K
bool
default y
select ARCH_32BIT_OFF_T
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_HAS_BINFMT_FLAT
select ARCH_HAS_CPU_FINALIZE_INIT if MMU
select ARCH_HAS_CURRENT_STACK_POINTER
new file mode 100644
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_M68K_CACHETYPE_H
+#define __ASM_M68K_CACHETYPE_H
+
+#include <linux/types.h>
+
+#define cpu_dcache_is_aliasing() true
+
+#endif
@@ -4,6 +4,7 @@ config MIPS
default y
select ARCH_32BIT_OFF_T if !64BIT
select ARCH_BINFMT_ELF_STATE if MIPS_FP_SUPPORT
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_HAS_CPU_FINALIZE_INIT
select ARCH_HAS_CURRENT_STACK_POINTER if !CC_IS_CLANG || CLANG_VERSION >= 140000
select ARCH_HAS_DEBUG_VIRTUAL if !64BIT
new file mode 100644
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_MIPS_CACHETYPE_H
+#define __ASM_MIPS_CACHETYPE_H
+
+#include <asm/cpu-features.h>
+
+#define cpu_dcache_is_aliasing() cpu_has_dc_aliases
+
+#endif
@@ -2,6 +2,7 @@
config NIOS2
def_bool y
select ARCH_32BIT_OFF_T
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_HAS_DMA_PREP_COHERENT
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
new file mode 100644
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_NIOS2_CACHETYPE_H
+#define __ASM_NIOS2_CACHETYPE_H
+
+#include <asm/page.h>
+#include <asm/cache.h>
+
+#define cpu_dcache_is_aliasing() (NIOS2_DCACHE_SIZE > PAGE_SIZE)
+
+#endif
@@ -8,6 +8,7 @@ config PARISC
select HAVE_FUNCTION_GRAPH_TRACER
select HAVE_SYSCALL_TRACEPOINTS
select ARCH_WANT_FRAME_POINTERS
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_HAS_DMA_ALLOC if PA11
select ARCH_HAS_ELF_RANDOMIZE
select ARCH_HAS_STRICT_KERNEL_RWX
new file mode 100644
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_PARISC_CACHETYPE_H
+#define __ASM_PARISC_CACHETYPE_H
+
+#include <linux/types.h>
+
+#define cpu_dcache_is_aliasing() true
+
+#endif
@@ -2,6 +2,7 @@
config SUPERH
def_bool y
select ARCH_32BIT_OFF_T
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_ENABLE_MEMORY_HOTPLUG if SPARSEMEM && MMU
select ARCH_ENABLE_MEMORY_HOTREMOVE if SPARSEMEM && MMU
select ARCH_HAVE_NMI_SAFE_CMPXCHG if (GUSA_RB || CPU_SH4A)
new file mode 100644
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_SH_CACHETYPE_H
+#define __ASM_SH_CACHETYPE_H
+
+#include <linux/types.h>
+
+#define cpu_dcache_is_aliasing() true
+
+#endif
@@ -13,6 +13,7 @@ config 64BIT
config SPARC
bool
default y
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_MIGHT_HAVE_PC_PARPORT if SPARC64 && PCI
select ARCH_MIGHT_HAVE_PC_SERIO
select DMA_OPS
new file mode 100644
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_SPARC_CACHETYPE_H
+#define __ASM_SPARC_CACHETYPE_H
+
+#include <asm/page.h>
+
+#ifdef CONFIG_SPARC32
+extern int vac_cache_size;
+#define cpu_dcache_is_aliasing() (vac_cache_size > PAGE_SIZE)
+#else
+#define cpu_dcache_is_aliasing() (L1DCACHE_SIZE > PAGE_SIZE)
+#endif
+
+#endif
@@ -2,6 +2,7 @@
config XTENSA
def_bool y
select ARCH_32BIT_OFF_T
+ select ARCH_HAS_CPU_CACHE_ALIASING
select ARCH_HAS_BINFMT_FLAT if !MMU
select ARCH_HAS_CURRENT_STACK_POINTER
select ARCH_HAS_DEBUG_VM_PGTABLE
new file mode 100644
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_XTENSA_CACHETYPE_H
+#define __ASM_XTENSA_CACHETYPE_H
+
+#include <asm/cache.h>
+#include <asm/page.h>
+
+#define cpu_dcache_is_aliasing() (DCACHE_WAY_SIZE > PAGE_SIZE)
+
+#endif
@@ -138,4 +138,10 @@ static inline int get_cpu_cacheinfo_id(int cpu, int level)
#define use_arch_cache_info() (false)
#endif
+#ifndef CONFIG_ARCH_HAS_CPU_CACHE_ALIASING
+#define cpu_dcache_is_aliasing() false
+#else
+#include <asm/cachetype.h>
+#endif
+
#endif /* _LINUX_CACHEINFO_H */
@@ -1016,6 +1016,12 @@ config IDLE_PAGE_TRACKING
See Documentation/admin-guide/mm/idle_page_tracking.rst for
more details.
+# Architectures which implement cpu_dcache_is_aliasing() to query
+# whether the data caches are aliased (VIVT or VIPT with dcache
+# aliasing) need to select this.
+config ARCH_HAS_CPU_CACHE_ALIASING
+ bool
+
config ARCH_HAS_CACHE_LINE_SIZE
bool