The patch titled
Subject: powerpc/64s/radix: enable huge vmalloc mappings
has been added to the -mm tree. Its filename is
powerpc-64s-radix-enable-huge-vmalloc-mappings.patch
This patch should soon appear at
https://ozlabs.org/~akpm/mmots/broken-out/powerpc-64s-radix-enable-huge-vmalloc-mappings.patch
and later at
https://ozlabs.org/~akpm/mmotm/broken-out/powerpc-64s-radix-enable-huge-vmalloc-mappings.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Nicholas Piggin <npiggin@xxxxxxxxx>
Subject: powerpc/64s/radix: enable huge vmalloc mappings
This reduces TLB misses by nearly 30x on a `git diff` workload on a 2-node
POWER9 (59,800 -> 2,100) and reduces CPU cycles by 0.54%, due to vfs
hashes being allocated with 2MB pages.
Link: https://lkml.kernel.org/r/20210317062402.533919-15-npiggin@xxxxxxxxx
Signed-off-by: Nicholas Piggin <npiggin@xxxxxxxxx>
Acked-by: Michael Ellerman <mpe@xxxxxxxxxxxxxx>
Cc: Borislav Petkov <bp@xxxxxxxxx>
Cc: Catalin Marinas <catalin.marinas@xxxxxxx>
Cc: Christoph Hellwig <hch@xxxxxx>
Cc: Ding Tianhong <dingtianhong@xxxxxxxxxx>
Cc: "H. Peter Anvin" <hpa@xxxxxxxxx>
Cc: Ingo Molnar <mingo@xxxxxxxxxx>
Cc: Miaohe Lin <linmiaohe@xxxxxxxxxx>
Cc: Russell King <linux@xxxxxxxxxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Uladzislau Rezki (Sony) <urezki@xxxxxxxxx>
Cc: Will Deacon <will@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
Documentation/admin-guide/kernel-parameters.txt | 2 +
arch/powerpc/Kconfig | 1
arch/powerpc/kernel/module.c | 22 +++++++++++---
3 files changed, 21 insertions(+), 4 deletions(-)
--- a/arch/powerpc/Kconfig~powerpc-64s-radix-enable-huge-vmalloc-mappings
+++ a/arch/powerpc/Kconfig
@@ -181,6 +181,7 @@ config PPC
select GENERIC_GETTIMEOFDAY
select HAVE_ARCH_AUDITSYSCALL
select HAVE_ARCH_HUGE_VMAP if PPC_BOOK3S_64 && PPC_RADIX_MMU
+ select HAVE_ARCH_HUGE_VMALLOC if HAVE_ARCH_HUGE_VMAP
select HAVE_ARCH_JUMP_LABEL
select HAVE_ARCH_KASAN if PPC32 && PPC_PAGE_SHIFT <= 14
select HAVE_ARCH_KASAN_VMALLOC if PPC32 && PPC_PAGE_SHIFT <= 14
--- a/arch/powerpc/kernel/module.c~powerpc-64s-radix-enable-huge-vmalloc-mappings
+++ a/arch/powerpc/kernel/module.c
@@ -8,6 +8,7 @@
#include <linux/moduleloader.h>
#include <linux/err.h>
#include <linux/vmalloc.h>
+#include <linux/mm.h>
#include <linux/bug.h>
#include <asm/module.h>
#include <linux/uaccess.h>
@@ -87,13 +88,26 @@ int module_finalize(const Elf_Ehdr *hdr,
return 0;
}
-#ifdef MODULES_VADDR
void *module_alloc(unsigned long size)
{
+ unsigned long start = VMALLOC_START;
+ unsigned long end = VMALLOC_END;
+
+#ifdef MODULES_VADDR
BUILD_BUG_ON(TASK_SIZE > MODULES_VADDR);
+ start = MODULES_VADDR;
+ end = MODULES_END;
+#endif
- return __vmalloc_node_range(size, 1, MODULES_VADDR, MODULES_END, GFP_KERNEL,
- PAGE_KERNEL_EXEC, VM_FLUSH_RESET_PERMS, NUMA_NO_NODE,
+ /*
+ * Don't do huge page allocations for modules yet until more testing
+ * is done. STRICT_MODULE_RWX may require extra work to support this
+ * too.
+ */
+
+ return __vmalloc_node_range(size, 1, start, end, GFP_KERNEL,
+ PAGE_KERNEL_EXEC,
+ VM_NO_HUGE_VMAP | VM_FLUSH_RESET_PERMS,
+ NUMA_NO_NODE,
__builtin_return_address(0));
}
-#endif
--- a/Documentation/admin-guide/kernel-parameters.txt~powerpc-64s-radix-enable-huge-vmalloc-mappings
+++ a/Documentation/admin-guide/kernel-parameters.txt
@@ -3243,6 +3243,8 @@
nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings.
+ nohugevmalloc [PPC] Disable kernel huge vmalloc mappings.
+
nosmt [KNL,S390] Disable symmetric multithreading (SMT).
Equivalent to smt=1.
_
Patches currently in -mm which might be from npiggin@xxxxxxxxx are
arm-mm-add-missing-pud_page-define-to-2-level-page-tables.patch
mm-vmalloc-fix-huge_vmap-regression-by-enabling-huge-pages-in-vmalloc_to_page.patch
mm-apply_to_pte_range-warn-and-fail-if-a-large-pte-is-encountered.patch
mm-vmalloc-rename-vmap__range-vmap_pages__range.patch
mm-ioremap-rename-ioremap__range-to-vmap__range.patch
mm-huge_vmap-arch-support-cleanup.patch
powerpc-inline-huge-vmap-supported-functions.patch
arm64-inline-huge-vmap-supported-functions.patch
x86-inline-huge-vmap-supported-functions.patch
mm-vmalloc-provide-fallback-arch-huge-vmap-support-functions.patch
mm-move-vmap_range-from-mm-ioremapc-to-mm-vmallocc.patch
mm-vmalloc-add-vmap_range_noflush-variant.patch
mm-vmalloc-hugepage-vmalloc-mappings.patch
powerpc-64s-radix-enable-huge-vmalloc-mappings.patch
