Skip to contentSkip to conversation

mm: Move ioremap page table mapping function to mm/

PATCH In review Nicholas Piggin 5 participants 24 messages Jun 10, 2019 – Jun 22, 2019

Subsystems (6)
lore
Share & tools
lore

Nicholas Piggin

author reviewer
#0 lore raw

PATCH 1/4 mm: Move ioremap page table mapping function to mm/

ioremap_page_range is a generic function to create a kernel virtual mapping, move it to mm/vmalloc.c and rename it vmap_range.

For clarity with this move, also:
- Rename vunmap_page_range (vmap_range's inverse) to vunmap_range.
- Rename vmap_page_range (which takes a page array) to vmap_pages.

---

Fixed up the arm64 compile errors, fixed a few bugs, and tidied things up a bit more.

Have tested powerpc and x86 but not arm64, would appreciate a review and test of the arm64 patch if possible.

 include/linux/vmalloc.h |   3 +
 lib/ioremap.c           | 173 +++---------------------------
 mm/vmalloc.c            | 228 ++++++++++++++++++++++++++++++++++++----
 3 files changed, 229 insertions(+), 175 deletions(-)
include/linux/vmalloc.h +3 -0
index 51e131245379..812bea5866d6 100644
@@ -147,6 +147,9 @@ extern struct vm_struct *find_vm_area(const void *addr);
 extern int map_vm_area(struct vm_struct *area, pgprot_t prot,
 			struct page **pages);
 #ifdef CONFIG_MMU
+extern int vmap_range(unsigned long addr,
+		       unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+		       unsigned int max_page_shift);
 extern int map_kernel_range_noflush(unsigned long start, unsigned long size,
 				    pgprot_t prot, struct page **pages);
 extern void unmap_kernel_range_noflush(unsigned long addr, unsigned long size);
lib/ioremap.c +16 -157

181 lines not on this page · raw message

mm/vmalloc.c +210 -18

338 lines not on this page · raw message

Signature · 1 line
2.20.1
author reviewer
#1 lore raw

PATCH 2/4 arm64: support huge vmap vmalloc

Applying huge vmap to vmalloc requires vmalloc_to_page to walk huge pages. Define pud_large and pmd_large to support this.

---
 arch/arm64/include/asm/pgtable.h | 2 ++
 1 file changed, 2 insertions(+)
arch/arm64/include/asm/pgtable.h +2 -0
index 2c41b04708fe..30fe7b344bf7 100644
@@ -428,6 +428,7 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 				 PMD_TYPE_TABLE)
 #define pmd_sect(pmd)		((pmd_val(pmd) & PMD_TYPE_MASK) == \
 				 PMD_TYPE_SECT)
+#define pmd_large(pmd)		pmd_sect(pmd)
 
 #if defined(CONFIG_ARM64_64K_PAGES) || CONFIG_PGTABLE_LEVELS < 3
 #define pud_sect(pud)		(0)
@@ -438,6 +439,7 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 #define pud_table(pud)		((pud_val(pud) & PUD_TYPE_MASK) == \
 				 PUD_TYPE_TABLE)
 #endif
+#define pud_large(pud)		pud_sect(pud)
 
 extern pgd_t init_pg_dir[PTRS_PER_PGD];
 extern pgd_t init_pg_end[];
Signature · 1 line
2.20.1
#2 lore raw

Re: [PATCH 2/4] arm64: support huge vmap vmalloc

Quoted 21 lines · Nicholas Piggin

Applying huge vmap to vmalloc requires vmalloc_to_page to walk huge pages. Define pud_large and pmd_large to support this.

---
 arch/arm64/include/asm/pgtable.h | 2 ++
 1 file changed, 2 insertions(+)
arch/arm64/include/asm/pgtable.h +1 -0
index 2c41b04708fe..30fe7b344bf7 100644
@@ -428,6 +428,7 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 				 PMD_TYPE_TABLE)
 #define pmd_sect(pmd)		((pmd_val(pmd) & PMD_TYPE_MASK) == \
 				 PMD_TYPE_SECT)
+#define pmd_large(pmd)		pmd_sect(pmd)
 
 #if defined(CONFIG_ARM64_64K_PAGES) || CONFIG_PGTABLE_LEVELS < 3
 #define pud_sect(pud)		(0)

arch/arm64/include/asm/pgtable.h extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, · line 445

Quoted 7 lines
arch/arm64/include/asm/pgtable.h +1 -0
@@ -438,6 +439,7 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 #define pud_table(pud)		((pud_val(pud) & PUD_TYPE_MASK) == \
 				 PUD_TYPE_TABLE)
 #endif
+#define pud_large(pud)		pud_sect(pud)
 
 extern pgd_t init_pg_dir[PTRS_PER_PGD];
 extern pgd_t init_pg_end[];

Another series (I guess not merged yet) is trying to add these wrappers on arm64 (https://patchwork.kernel.org/patch/10883887/).

author reviewer
#3 lore raw

Re: [PATCH 2/4] arm64: support huge vmap vmalloc

Anshuman Khandual's on June 10, 2019 3:47 pm:

Quoted 29 lines · Nicholas Piggin
Quoted 26 lines

Applying huge vmap to vmalloc requires vmalloc_to_page to walk huge pages. Define pud_large and pmd_large to support this.

---
 arch/arm64/include/asm/pgtable.h | 2 ++
 1 file changed, 2 insertions(+)
arch/arm64/include/asm/pgtable.h +2 -0
index 2c41b04708fe..30fe7b344bf7 100644
@@ -428,6 +428,7 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 				 PMD_TYPE_TABLE)
 #define pmd_sect(pmd)		((pmd_val(pmd) & PMD_TYPE_MASK) == \
 				 PMD_TYPE_SECT)
+#define pmd_large(pmd)		pmd_sect(pmd)
 
 #if defined(CONFIG_ARM64_64K_PAGES) || CONFIG_PGTABLE_LEVELS < 3
 #define pud_sect(pud)		(0)
@@ -438,6 +439,7 @@ extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn,
 #define pud_table(pud)		((pud_val(pud) & PUD_TYPE_MASK) == \
 				 PUD_TYPE_TABLE)
 #endif
+#define pud_large(pud)		pud_sect(pud)
 
Quoted 6 lines

arch/arm64/include/asm/pgtable.h extern pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, · line 445

Quoted 2 lines
arch/arm64/include/asm/pgtable.h +0 -0
 extern pgd_t init_pg_dir[PTRS_PER_PGD];
 extern pgd_t init_pg_end[];

Another series (I guess not merged yet) is trying to add these wrappers on arm64 (https://patchwork.kernel.org/patch/10883887/).

Okay good, I'll just cherry pick it for the series.

Thanks,
Nick

author reviewer
#4 lore raw

PATCH 3/4 powerpc/64s/radix: support huge vmap vmalloc

In reply to Nicholas Piggin

Applying huge vmap to vmalloc requires vmalloc_to_page to walk huge pages. Define pud_large and pmd_large to support this.

---
 arch/powerpc/include/asm/book3s/64/pgtable.h | 24 ++++++++++++--------
 1 file changed, 15 insertions(+), 9 deletions(-)
arch/powerpc/include/asm/book3s/64/pgtable.h +15 -9
index 5faceeefd9f9..8e02077b11fb 100644
@@ -923,6 +923,11 @@ static inline int pud_present(pud_t pud)
 	return !!(pud_raw(pud) & cpu_to_be64(_PAGE_PRESENT));
 }
 
+static inline int pud_large(pud_t pud)
+{
+	return !!(pud_raw(pud) & cpu_to_be64(_PAGE_PTE));
+}
+
 extern struct page *pud_page(pud_t pud);
 extern struct page *pmd_page(pmd_t pmd);
 static inline pte_t pud_pte(pud_t pud)
@@ -966,6 +971,11 @@ static inline int pgd_present(pgd_t pgd)
 	return !!(pgd_raw(pgd) & cpu_to_be64(_PAGE_PRESENT));
 }
 
+static inline int pgd_large(pgd_t pgd)
+{
+	return !!(pgd_raw(pgd) & cpu_to_be64(_PAGE_PTE));
+}
+
 static inline pte_t pgd_pte(pgd_t pgd)
 {
 	return __pte_raw(pgd_raw(pgd));
@@ -1091,6 +1101,11 @@ static inline pte_t *pmdp_ptep(pmd_t *pmd)
 #define pmd_mk_savedwrite(pmd)	pte_pmd(pte_mk_savedwrite(pmd_pte(pmd)))
 #define pmd_clear_savedwrite(pmd)	pte_pmd(pte_clear_savedwrite(pmd_pte(pmd)))
 
+static inline int pmd_large(pmd_t pmd)
+{
+	return !!(pmd_raw(pmd) & cpu_to_be64(_PAGE_PTE));
+}
+
 #ifdef CONFIG_HAVE_ARCH_SOFT_DIRTY
 #define pmd_soft_dirty(pmd)    pte_soft_dirty(pmd_pte(pmd))
 #define pmd_mksoft_dirty(pmd)  pte_pmd(pte_mksoft_dirty(pmd_pte(pmd)))
@@ -1159,15 +1174,6 @@ pmd_hugepage_update(struct mm_struct *mm, unsigned long addr, pmd_t *pmdp,
 	return hash__pmd_hugepage_update(mm, addr, pmdp, clr, set);
 }
 
-/*
- * returns true for pmd migration entries, THP, devmap, hugetlb
- * But compile time dependent on THP config
- */
-static inline int pmd_large(pmd_t pmd)
-{
-	return !!(pmd_raw(pmd) & cpu_to_be64(_PAGE_PTE));
-}
-
 static inline pmd_t pmd_mknotpresent(pmd_t pmd)
 {
 	return __pmd(pmd_val(pmd) & ~_PAGE_PRESENT);
Signature · 1 line
2.20.1
author reviewer
#5 lore raw

PATCH 4/4 mm/vmalloc: Hugepage vmalloc mappings

In reply to Nicholas Piggin

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

---
 include/asm-generic/4level-fixup.h |   1 +
 include/asm-generic/5level-fixup.h |   1 +
 include/linux/vmalloc.h            |   1 +
 mm/vmalloc.c                       | 132 +++++++++++++++++++++++------
 4 files changed, 107 insertions(+), 28 deletions(-)
include/asm-generic/4level-fixup.h +1 -0
index e3667c9a33a5..3cc65a4dd093 100644
@@ -20,6 +20,7 @@
 #define pud_none(pud)			0
 #define pud_bad(pud)			0
 #define pud_present(pud)		1
+#define pud_large(pud)			0
 #define pud_ERROR(pud)			do { } while (0)
 #define pud_clear(pud)			pgd_clear(pud)
 #define pud_val(pud)			pgd_val(pud)
include/asm-generic/5level-fixup.h +1 -0
index bb6cb347018c..c4377db09a4f 100644
@@ -22,6 +22,7 @@
 #define p4d_none(p4d)			0
 #define p4d_bad(p4d)			0
 #define p4d_present(p4d)		1
+#define p4d_large(p4d)			0
 #define p4d_ERROR(p4d)			do { } while (0)
 #define p4d_clear(p4d)			pgd_clear(p4d)
 #define p4d_val(p4d)			pgd_val(p4d)
include/linux/vmalloc.h +1 -0
index 812bea5866d6..4c92dc608928 100644
@@ -42,6 +42,7 @@ struct vm_struct {
 	unsigned long		size;
 	unsigned long		flags;
 	struct page		**pages;
+	unsigned int		page_shift;
 	unsigned int		nr_pages;
 	phys_addr_t		phys_addr;
 	const void		*caller;
mm/vmalloc.c +104 -28

233 lines not on this page · raw message

Signature · 1 line
2.20.1
author reviewer
#6 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Nicholas Piggin's on June 10, 2019 2:38 pm:

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 475

Quoted 7 lines
mm/vmalloc.c +7 -0
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+			   pgprot_t prot, struct page **pages,
+			   unsigned int page_shift)
+{
+	BUG_ON(page_shift != PAGE_SIZE);
+	return vmap_pages_range(start, end, prot, pages);
+}

That's a false positive BUG_ON for !HUGE_VMAP configs. I'll fix that and repost after a round of feedback.

Thanks,
Nick

#7 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Quoted 11 lines · Nicholas Piggin

Nicholas Piggin's on June 10, 2019 2:38 pm:

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 475

Quoted 7 lines
mm/vmalloc.c +7 -0
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+			   pgprot_t prot, struct page **pages,
+			   unsigned int page_shift)
+{
+	BUG_ON(page_shift != PAGE_SIZE);
+	return vmap_pages_range(start, end, prot, pages);
+}

That's a false positive BUG_ON for !HUGE_VMAP configs. I'll fix that and repost after a round of feedback.

Sure, Crash log for that false positive BUG_ON on Power8 Host.

[    0.001718] pid_max: default: 163840 minimum: 1280
[    0.010437] ------------[ cut here ]------------
[    0.010461] kernel BUG at mm/vmalloc.c:473!
[    0.010471] Oops: Exception in kernel mode, sig: 5 [#1]
[    0.010481] LE PAGE_SIZE=64K MMU=Hash SMP NR_CPUS=2048 NUMA PowerNV
[    0.010491] Modules linked in:
[    0.010503] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.2.0-rc3-ga7ee9421d #1
[    0.010515] NIP:  c00000000034dbd8 LR: c00000000034dc80 CTR: 0000000000000000
[    0.010527] REGS: c0000000015bf9a0 TRAP: 0700   Not tainted  (5.2.0-rc3-ga7ee9421d)
[    0.010537] MSR:  9000000002029033 <SF,HV,VEC,EE,ME,IR,DR,RI,LE>  CR: 22022422  XER: 20000000
[    0.010559] CFAR: c00000000034dc88 IRQMASK: 0
[    0.010559] GPR00: c00000000034dc80 c0000000015bfc30 c0000000015c2f00 c00c000001fd0e00
[    0.010559] GPR04: 0000000000000000 0000000000002322 0000000000000000 0000000000000040
[    0.010559] GPR08: c000000ff9080000 0000000000000400 0000000000000400 0000000000000100
[    0.010559] GPR12: 0000000042022422 c0000000017a0000 00000001035ae7d8 0000000000000400
[    0.010559] GPR16: 0000000004000000 800000000000018e c000000000ee08c8 0000000000000000
[    0.010559] GPR20: 0000000000010000 0000000000002b22 0000000000000b20 0000000000000022
[    0.010559] GPR24: c0000007f92c7880 0000000000000b22 0000000000010000 c00a000000000000
[    0.010559] GPR28: c008000000000000 0000000004000000 ffffffffffffffff 0000000000000b20
[    0.010664] NIP [c00000000034dbd8] __vmalloc_node_range+0x1f8/0x410
[    0.010677] LR [c00000000034dc80] __vmalloc_node_range+0x2a0/0x410
[    0.010686] Call Trace:
[    0.010695] [c0000000015bfc30] [c00000000034dc80] __vmalloc_node_range+0x2a0/0x410 (unreliable)
[    0.010711] [c0000000015bfd30] [c00000000034de40] __vmalloc+0x50/0x60
[    0.010724] [c0000000015bfda0] [c00000000101e54c] alloc_large_system_hash+0x200/0x304
[    0.010738] [c0000000015bfe60] [c0000000010235bc] vfs_caches_init+0xd8/0x138
[    0.010752] [c0000000015bfee0] [c000000000fe428c] start_kernel+0x5c4/0x668
[    0.010767] [c0000000015bff90] [c00000000000b774] start_here_common+0x1c/0x528
[    0.010777] Instruction dump:
[    0.010785] 60000000 7c691b79 418200dc e9180020 79ea1f24 7d28512a 40920170 8138002c
[    0.010803] 394f0001 794f0020 7c095040 4181ffbc <0fe00000> 60000000 3f400001 4bfffedc
[    0.010826] ---[ end trace dd0217488686d653 ]---
[    0.010834]
[    1.010946] Kernel panic - not syncing: Attempted to kill the idle task!
[    1.011061] Rebooting in 10 seconds..

Regards,
-Satheesh.

Quoted 5 lines

Thanks,
Nick

#8 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

In reply to Nicholas Piggin
Quoted 2 lines · Nicholas Piggin

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them.

IIUC that extends HAVE_ARCH_HUGE_VMAP from iormap to vmalloc.

Quoted 5 lines

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

Sure will try this on arm64.

Quoted 26 lines
---
 include/asm-generic/4level-fixup.h |   1 +
 include/asm-generic/5level-fixup.h |   1 +
 include/linux/vmalloc.h            |   1 +
 mm/vmalloc.c                       | 132 +++++++++++++++++++++++------
 4 files changed, 107 insertions(+), 28 deletions(-)
include/asm-generic/4level-fixup.h +1 -0
index e3667c9a33a5..3cc65a4dd093 100644
@@ -20,6 +20,7 @@
 #define pud_none(pud)			0
 #define pud_bad(pud)			0
 #define pud_present(pud)		1
+#define pud_large(pud)			0
 #define pud_ERROR(pud)			do { } while (0)
 #define pud_clear(pud)			pgd_clear(pud)
 #define pud_val(pud)			pgd_val(pud)
include/asm-generic/5level-fixup.h +0 -0
index bb6cb347018c..c4377db09a4f 100644

include/asm-generic/5level-fixup.h line 28

Quoted 7 lines
include/asm-generic/5level-fixup.h +1 -0
@@ -22,6 +22,7 @@
 #define p4d_none(p4d)			0
 #define p4d_bad(p4d)			0
 #define p4d_present(p4d)		1
+#define p4d_large(p4d)			0
 #define p4d_ERROR(p4d)			do { } while (0)
 #define p4d_clear(p4d)			pgd_clear(p4d)
 #define p4d_val(p4d)			pgd_val(p4d)

Both of these are required from vmalloc_to_page() which as per a later comment should be part of a prerequisite patch before this series.

include/linux/vmalloc.h struct vm_struct { · line 45

Quoted 9 lines
include/linux/vmalloc.h +1 -0
index 812bea5866d6..4c92dc608928 100644
@@ -42,6 +42,7 @@ struct vm_struct {
 	unsigned long		size;
 	unsigned long		flags;
 	struct page		**pages;
+	unsigned int		page_shift;

So the entire vm_struct will be mapped with a single page_shift. It cannot have mix and match mappings with PAGE_SIZE, PMD_SIZE, PUD_SIZE etc in case the allocation fails for larger ones, falling back etc what over other reasons.

Quoted 16 lines
include/linux/vmalloc.h +0 -0
 	unsigned int		nr_pages;
 	phys_addr_t		phys_addr;
 	const void		*caller;
mm/vmalloc.c +1 -0
index dd27cfb29b10..0cf8e861caeb 100644
@@ -36,6 +36,7 @@
 #include <linux/rbtree_augmented.h>
 
 #include <linux/uaccess.h>
+#include <asm/pgtable.h>
 #include <asm/tlbflush.h>
 #include <asm/shmparam.h>
 

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 445

Quoted 5 lines
mm/vmalloc.c +2 -0
@@ -440,6 +441,41 @@ static int vmap_pages_range(unsigned long start, unsigned long end,
 	return ret;
 }
 
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+static int vmap_hpages_range(unsigned long start, unsigned long end,

A small nit (if you agree) s/hpages/huge_pages/

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 446

Quoted 1 line
mm/vmalloc.c +1 -0
+				   pgprot_t prot, struct page **pages,

Re-order (prot <---> pages) just to follow the standard like before.

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 450

Quoted 4 lines
mm/vmalloc.c +4 -0
+				   unsigned int page_shift)
+{
+	unsigned long addr = start;
+	unsigned int i, nr = (end - start) >> (PAGE_SHIFT + page_shift);

s/nr/nr_huge_pages ?

Also should not we check for the alignment of the range [start...end] with respect to (1UL << [PAGE_SHIFT + page_shift]).

Quoted 41 lines
mm/vmalloc.c +29 -1
+
+	for (i = 0; i < nr; i++) {
+		int err;
+
+		err = vmap_range_noflush(addr,
+					addr + (PAGE_SIZE << page_shift),
+					__pa(page_address(pages[i])), prot,
+					PAGE_SHIFT + page_shift);
+		if (err)
+			return err;
+
+		addr += PAGE_SIZE << page_shift;
+	}
+	flush_cache_vmap(start, end);
+
+	return nr;
+}
+#else
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+			   pgprot_t prot, struct page **pages,
+			   unsigned int page_shift)
+{
+	BUG_ON(page_shift != PAGE_SIZE);
+	return vmap_pages_range(start, end, prot, pages);
+}
+#endif
+
+
 int is_vmalloc_or_module_addr(const void *x)
 {
 	/*
@@ -462,7 +498,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 {
 	unsigned long addr = (unsigned long) vmalloc_addr;
 	struct page *page = NULL;
-	pgd_t *pgd = pgd_offset_k(addr);
+	pgd_t *pgd;
 	p4d_t *p4d;
 	pud_t *pud;
 	pmd_t *pmd;

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 516

Quoted 7 lines
mm/vmalloc.c +2 -0
@@ -474,27 +510,38 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	 */
 	VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr));
 
+	pgd = pgd_offset_k(addr);
 	if (pgd_none(*pgd))
 		return NULL;
+

Small nit. Stray line here.

'pgd' related changes here seem to be just cleanups and should not part of this patch.

Quoted 9 lines
mm/vmalloc.c +5 -1
 	p4d = p4d_offset(pgd, addr);
 	if (p4d_none(*p4d))
 		return NULL;
-	pud = pud_offset(p4d, addr);
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (p4d_large(*p4d))
+		return p4d_page(*p4d) + ((addr & ~P4D_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(p4d_bad(*p4d)))

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 491

Quoted 10 lines
mm/vmalloc.c +1 -8
+		return NULL;
 
-	/*
-	 * Don't dereference bad PUD or PMD (below) entries. This will also
-	 * identify huge mappings, which we may encounter on architectures
-	 * that define CONFIG_HAVE_ARCH_HUGE_VMAP=y. Such regions will be
-	 * identified as vmalloc addresses by is_vmalloc_addr(), but are
-	 * not [unambiguously] associated with a struct page, so there is
-	 * no correct value to return for them.
-	 */

What changed the situation so that we could return struct page for a huge mapping now ? AFAICT even after this patch, PUD/P4D level huge pages can only be created with ioremap_page_range() not with vmalloc() which creates PMD sized mappings only. Hence if it's okay to dereference struct page of a huge mapping (not withstanding the comment here) it should be part of an earlier patch fixing it first for existing ioremap_page_range() huge mappings.

Quoted 13 lines
mm/vmalloc.c +9 -2
-	WARN_ON_ONCE(pud_bad(*pud));
-	if (pud_none(*pud) || pud_bad(*pud))
+	pud = pud_offset(p4d, addr);
+	if (pud_none(*pud))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pud_large(*pud))
+		return pud_page(*pud) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pud_bad(*pud)))
 		return NULL;
+
 	pmd = pmd_offset(pud, addr);

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 545

Quoted 10 lines
mm/vmalloc.c +7 -2
-	WARN_ON_ONCE(pmd_bad(*pmd));
-	if (pmd_none(*pmd) || pmd_bad(*pmd))
+	if (pmd_none(*pmd))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pmd_large(*pmd))
+		return pmd_page(*pmd) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pmd_bad(*pmd)))
 		return NULL;

At each page table level, we are checking in this order

pXX_none() --> pXX_large() --> pXX_bad()

Are not these alternative orders bit better

pXX_bad() --> pXX_none() --> pXX_large()

Or

pXX_none() --> pXX_bad() --> pXX_large()

Checking for pXX_bad() at the end does not make much sense.

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 552

Quoted 7 lines
mm/vmalloc.c +1 -0
 	ptep = pte_offset_map(pmd, addr);
@@ -502,6 +549,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	if (pte_present(pte))
 		page = pte_page(pte);
 	pte_unmap(ptep);
+

Small nit. Stray line here.

Quoted 16 lines
mm/vmalloc.c +3 -2
 	return page;
 }
 EXPORT_SYMBOL(vmalloc_to_page);
@@ -2185,8 +2233,9 @@ static struct vm_struct *__get_vm_area_node(unsigned long size,
 		return NULL;
 
 	if (flags & VM_IOREMAP)
-		align = 1ul << clamp_t(int, get_count_order_long(size),
-				       PAGE_SHIFT, IOREMAP_MAX_ORDER);
+		align = max(align,
+				1ul << clamp_t(int, get_count_order_long(size),
+				       PAGE_SHIFT, IOREMAP_MAX_ORDER));
 
 	area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node);
 	if (unlikely(!area))

mm/vmalloc.c static void __vunmap(const void *addr, int deallocate_pages) · line 2450

Quoted 5 lines
mm/vmalloc.c +1 -1
@@ -2398,7 +2447,7 @@ static void __vunmap(const void *addr, int deallocate_pages)
 			struct page *page = area->pages[i];
 
 			BUG_ON(!page);
-			__free_pages(page, 0);
+			__free_pages(page, area->page_shift);

area->page_shift' turns out to be effective page order. I think the name here is bit misleading. s/page_shift/page_order or nr_pages should be better IMHO. page_shift is not actual shift (1UL << area->shift to get size) nor does it sound like page 'order'.

Quoted 25 lines
mm/vmalloc.c +6 -3
 		}
 
 		kvfree(area->pages);
@@ -2541,14 +2590,17 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 				 pgprot_t prot, int node)
 {
 	struct page **pages;
+	unsigned long addr = (unsigned long)area->addr;
+	unsigned long size = get_vm_area_size(area);
+	unsigned int page_shift = area->page_shift;
+	unsigned int shift = page_shift + PAGE_SHIFT;
 	unsigned int nr_pages, array_size, i;
 	const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO;
 	const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN;
 	const gfp_t highmem_mask = (gfp_mask & (GFP_DMA | GFP_DMA32)) ?
-					0 :
-					__GFP_HIGHMEM;
+					0 : __GFP_HIGHMEM;
 
-	nr_pages = get_vm_area_size(area) >> PAGE_SHIFT;
+	nr_pages = size >> shift;
 	array_size = (nr_pages * sizeof(struct page *));
 
 	area->nr_pages = nr_pages;

mm/vmalloc.c static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, · line 2625

Quoted 9 lines
mm/vmalloc.c +2 -4
@@ -2569,10 +2621,8 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 	for (i = 0; i < area->nr_pages; i++) {
 		struct page *page;
 
-		if (node == NUMA_NO_NODE)
-			page = alloc_page(alloc_mask|highmem_mask);
-		else
-			page = alloc_pages_node(node, alloc_mask|highmem_mask, 0);
+		page = alloc_pages_node(node,
+				alloc_mask|highmem_mask, page_shift);

alloc_mask remains the exact same like before even for these high order pages.

Quoted 27 lines
mm/vmalloc.c +5 -2
 		if (unlikely(!page)) {
 			/* Successfully allocated i pages, free them in __vunmap() */
@@ -2584,8 +2634,9 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 			cond_resched();
 	}
 
-	if (map_vm_area(area, prot, pages))
+	if (vmap_hpages_range(addr, addr + size, prot, pages, page_shift) < 0)
 		goto fail;
+
 	return area->addr;
 
 fail:
@@ -2619,22 +2670,39 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
 			pgprot_t prot, unsigned long vm_flags, int node,
 			const void *caller)
 {
-	struct vm_struct *area;
+	struct vm_struct *area = NULL;
 	void *addr;
 	unsigned long real_size = size;
+	unsigned long real_align = align;
+	unsigned int shift = PAGE_SHIFT;
 
 	size = PAGE_ALIGN(size);
 	if (!size || (size >> PAGE_SHIFT) > totalram_pages())

mm/vmalloc.c void *__vmalloc_node_range(unsigned long size, unsigned long align, · line 2690

Quoted 10 lines
mm/vmalloc.c +8 -0
 		goto fail;
 
+	if (IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP)) {
+		unsigned long size_per_node;
+
+		size_per_node = size;
+		if (node == NUMA_NO_NODE)
+			size_per_node /= num_online_nodes();
+		if (size_per_node >= PMD_SIZE)
+			shift = PMD_SHIFT;

There are two problems here.

1. Should not size_per_node be aligned with PMD_SIZE to avoid wasting memory later because of alignment upwards (making it worse for NUMA_NO_NODE)
2. What about PUD_SIZE which is not considered here at all
3. We should have similar knobs like ioremap controlling different size huge mappings

static int __read_mostly ioremap_p4d_capable;
static int __read_mostly ioremap_pud_capable;
static int __read_mostly ioremap_pmd_capable;
static int __read_mostly ioremap_huge_disabled;

while also giving arch a chance to weigh in through similar overrides like these.

arch_ioremap_[pud|pmd]_supported() ---> probably unifying it for vmalloc()

Quoted 20 lines
mm/vmalloc.c +8 -1
+	}
+again:
+	align = max(real_align, 1UL << shift);
+	size = ALIGN(real_size, align);
+
 	area = __get_vm_area_node(size, align, VM_ALLOC | VM_UNINITIALIZED |
 				vm_flags, start, end, node, gfp_mask, caller);
 	if (!area)
 		goto fail;
 
+	area->page_shift = shift - PAGE_SHIFT;
+
 	addr = __vmalloc_area_node(area, gfp_mask, prot, node);
 	if (!addr)
-		return NULL;
+		goto fail;
 
 	/*
 	 * In this function, newly allocated vm_struct has VM_UNINITIALIZED

mm/vmalloc.c void *__vmalloc_node_range(unsigned long size, unsigned long align, · line 2722

Quoted 8 lines
mm/vmalloc.c +4 -1
@@ -2648,8 +2716,16 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
 	return addr;
 
 fail:
-	warn_alloc(gfp_mask, NULL,
+	if (shift == PMD_SHIFT) {
+		shift = PAGE_SHIFT;
+		goto again;
+	}

PUD_SHIFT should be accommodated here as well while falling back to lower mapping sizes in case previous allocation attempt fails.

author reviewer
#9 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Anshuman Khandual's on June 10, 2019 6:53 pm:

Quoted 42 lines · Nicholas Piggin
Quoted 2 lines

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them.

IIUC that extends HAVE_ARCH_HUGE_VMAP from iormap to vmalloc.

Quoted 5 lines

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

Sure will try this on arm64.

Quoted 28 lines
---
 include/asm-generic/4level-fixup.h |   1 +
 include/asm-generic/5level-fixup.h |   1 +
 include/linux/vmalloc.h            |   1 +
 mm/vmalloc.c                       | 132 +++++++++++++++++++++++------
 4 files changed, 107 insertions(+), 28 deletions(-)
include/asm-generic/4level-fixup.h +1 -0
index e3667c9a33a5..3cc65a4dd093 100644
@@ -20,6 +20,7 @@
 #define pud_none(pud)			0
 #define pud_bad(pud)			0
 #define pud_present(pud)		1
+#define pud_large(pud)			0
 #define pud_ERROR(pud)			do { } while (0)
 #define pud_clear(pud)			pgd_clear(pud)
 #define pud_val(pud)			pgd_val(pud)
include/asm-generic/5level-fixup.h +0 -0
index bb6cb347018c..c4377db09a4f 100644
@@ -22,6 +22,7 @@
 #define p4d_none(p4d)			0
 #define p4d_bad(p4d)			0
Quoted 8 lines

include/asm-generic/5level-fixup.h line 28

Quoted 5 lines
include/asm-generic/5level-fixup.h +1 -0
 #define p4d_present(p4d)		1
+#define p4d_large(p4d)			0
 #define p4d_ERROR(p4d)			do { } while (0)
 #define p4d_clear(p4d)			pgd_clear(p4d)
 #define p4d_val(p4d)			pgd_val(p4d)

Both of these are required from vmalloc_to_page() which as per a later comment should be part of a prerequisite patch before this series.

I'm not sure what you mean. This patch is where they get used.

Possibly I could split this and the vmalloc_to_page change out. I'll consider it.

Quoted 5 lines
Quoted 5 lines
include/linux/vmalloc.h +0 -0
index 812bea5866d6..4c92dc608928 100644
@@ -42,6 +42,7 @@ struct vm_struct {
Quoted 8 lines

include/linux/vmalloc.h struct vm_struct { · line 45

Quoted 4 lines
include/linux/vmalloc.h +1 -0
 	unsigned long		size;
 	unsigned long		flags;
 	struct page		**pages;
+	unsigned int		page_shift;

So the entire vm_struct will be mapped with a single page_shift. It cannot have mix and match mappings with PAGE_SIZE, PMD_SIZE, PUD_SIZE etc in case the allocation fails for larger ones, falling back etc what over other reasons.

For now, yes. I have a bit of follow up work to improve that and make it able to fall back, but it's a bit more churn and not a significant benefit just yet because there are not a lot of very large vmallocs (except the early hashes which can be satisfied with large allocs).

Quoted 16 lines
Quoted 16 lines
include/linux/vmalloc.h +0 -0
 	unsigned int		nr_pages;
 	phys_addr_t		phys_addr;
 	const void		*caller;
mm/vmalloc.c +1 -0
index dd27cfb29b10..0cf8e861caeb 100644
@@ -36,6 +36,7 @@
 #include <linux/rbtree_augmented.h>
 
 #include <linux/uaccess.h>
+#include <asm/pgtable.h>
 #include <asm/tlbflush.h>
 #include <asm/shmparam.h>
 
Quoted 8 lines

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 445

Quoted 6 lines
mm/vmalloc.c +2 -0
@@ -440,6 +441,41 @@ static int vmap_pages_range(unsigned long start, unsigned long end,
 	return ret;
 }
 
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+static int vmap_hpages_range(unsigned long start, unsigned long end,

A small nit (if you agree) s/hpages/huge_pages/

Hmm. It's not actually a good function name because it can do small pages as well. vmap_pages_size_range or something may be better.

Quoted 4 lines

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 446

Quoted 2 lines
mm/vmalloc.c +1 -0
+				   pgprot_t prot, struct page **pages,

Re-order (prot <---> pages) just to follow the standard like before.

Will do.

Quoted 6 lines

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 450

Quoted 4 lines
mm/vmalloc.c +4 -0
+				   unsigned int page_shift)
+{
+	unsigned long addr = start;
+	unsigned int i, nr = (end - start) >> (PAGE_SHIFT + page_shift);

s/nr/nr_huge_pages ?

Sure.

Quoted 2 lines

Also should not we check for the alignment of the range [start...end] with respect to (1UL << [PAGE_SHIFT + page_shift]).

The caller should if it specifies large page. Could check and -EINVAL for incorrect alignment.

Quoted 44 lines
Quoted 44 lines
mm/vmalloc.c +29 -1
+
+	for (i = 0; i < nr; i++) {
+		int err;
+
+		err = vmap_range_noflush(addr,
+					addr + (PAGE_SIZE << page_shift),
+					__pa(page_address(pages[i])), prot,
+					PAGE_SHIFT + page_shift);
+		if (err)
+			return err;
+
+		addr += PAGE_SIZE << page_shift;
+	}
+	flush_cache_vmap(start, end);
+
+	return nr;
+}
+#else
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+			   pgprot_t prot, struct page **pages,
+			   unsigned int page_shift)
+{
+	BUG_ON(page_shift != PAGE_SIZE);
+	return vmap_pages_range(start, end, prot, pages);
+}
+#endif
+
+
 int is_vmalloc_or_module_addr(const void *x)
 {
 	/*
@@ -462,7 +498,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 {
 	unsigned long addr = (unsigned long) vmalloc_addr;
 	struct page *page = NULL;
-	pgd_t *pgd = pgd_offset_k(addr);
+	pgd_t *pgd;
 	p4d_t *p4d;
 	pud_t *pud;
 	pmd_t *pmd;
@@ -474,27 +510,38 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	 */
 	VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr));
 
Quoted 8 lines

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 516

Quoted 4 lines
mm/vmalloc.c +2 -0
+	pgd = pgd_offset_k(addr);
 	if (pgd_none(*pgd))
 		return NULL;
+

Small nit. Stray line here.

'pgd' related changes here seem to be just cleanups and should not part of this patch.

Yeah I figure it doesn't matter to make small changes close by, but maybe that's more frowned upon now for git blame?

Quoted 11 lines
Quoted 11 lines
mm/vmalloc.c +6 -1
 	p4d = p4d_offset(pgd, addr);
 	if (p4d_none(*p4d))
 		return NULL;
-	pud = pud_offset(p4d, addr);
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (p4d_large(*p4d))
+		return p4d_page(*p4d) + ((addr & ~P4D_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(p4d_bad(*p4d)))
+		return NULL;
 
Quoted 11 lines

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 491

Quoted 8 lines
mm/vmalloc.c +0 -8
-	/*
-	 * Don't dereference bad PUD or PMD (below) entries. This will also
-	 * identify huge mappings, which we may encounter on architectures
-	 * that define CONFIG_HAVE_ARCH_HUGE_VMAP=y. Such regions will be
-	 * identified as vmalloc addresses by is_vmalloc_addr(), but are
-	 * not [unambiguously] associated with a struct page, so there is
-	 * no correct value to return for them.
-	 */

What changed the situation so that we could return struct page for a huge mapping now ?

For the PUD case? Nothing changed, per se, we I just calculate the correct struct page now, so I may return it.

Quoted 5 lines

AFAICT even after this patch, PUD/P4D level huge pages can only be created with ioremap_page_range() not with vmalloc() which creates PMD sized mappings only. Hence if it's okay to dereference struct page of a huge mapping (not withstanding the comment here) it should be part of an earlier patch fixing it first for existing ioremap_page_range() huge mappings.

Possibly yes, we can consider 029c54b095995 to be a band-aid for huge vmaps which is fixed properly by this change, in which case it could make sense to break this into its own patch.

Quoted 31 lines
Quoted 14 lines
mm/vmalloc.c +9 -2
-	WARN_ON_ONCE(pud_bad(*pud));
-	if (pud_none(*pud) || pud_bad(*pud))
+	pud = pud_offset(p4d, addr);
+	if (pud_none(*pud))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pud_large(*pud))
+		return pud_page(*pud) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pud_bad(*pud)))
 		return NULL;
+
 	pmd = pmd_offset(pud, addr);

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 545

Quoted 10 lines
mm/vmalloc.c +7 -2
-	WARN_ON_ONCE(pmd_bad(*pmd));
-	if (pmd_none(*pmd) || pmd_bad(*pmd))
+	if (pmd_none(*pmd))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pmd_large(*pmd))
+		return pmd_page(*pmd) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pmd_bad(*pmd)))
 		return NULL;

At each page table level, we are checking in this order

pXX_none() --> pXX_large() --> pXX_bad()

Are not these alternative orders bit better

Quoted 7 lines

pXX_bad() --> pXX_none() --> pXX_large()

Or

pXX_none() --> pXX_bad() --> pXX_large()

Checking for pXX_bad() at the end does not make much sense.

Yeah the order tends to go none->bad. It's not 100% clear we can test bad before none (at least it goes against convention). But good point should be changed to your last sequence I think.

Quoted 10 lines

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 552

Quoted 8 lines
mm/vmalloc.c +1 -0
 	ptep = pte_offset_map(pmd, addr);
@@ -502,6 +549,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	if (pte_present(pte))
 		page = pte_page(pte);
 	pte_unmap(ptep);
+

Small nit. Stray line here.

I don't mind adding some lines here and there, like here. It is an unrelated (alleged-)cleanup though.

Quoted 19 lines
Quoted 19 lines
mm/vmalloc.c +3 -2
 	return page;
 }
 EXPORT_SYMBOL(vmalloc_to_page);
@@ -2185,8 +2233,9 @@ static struct vm_struct *__get_vm_area_node(unsigned long size,
 		return NULL;
 
 	if (flags & VM_IOREMAP)
-		align = 1ul << clamp_t(int, get_count_order_long(size),
-				       PAGE_SHIFT, IOREMAP_MAX_ORDER);
+		align = max(align,
+				1ul << clamp_t(int, get_count_order_long(size),
+				       PAGE_SHIFT, IOREMAP_MAX_ORDER));
 
 	area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node);
 	if (unlikely(!area))
@@ -2398,7 +2447,7 @@ static void __vunmap(const void *addr, int deallocate_pages)
 			struct page *page = area->pages[i];
 
Quoted 7 lines

mm/vmalloc.c static void __vunmap(const void *addr, int deallocate_pages) · line 2450

Quoted 3 lines
mm/vmalloc.c +1 -1
 			BUG_ON(!page);
-			__free_pages(page, 0);
+			__free_pages(page, area->page_shift);

area->page_shift' turns out to be effective page order. I think the name here is bit misleading. s/page_shift/page_order or nr_pages should be better IMHO. page_shift is not actual shift (1UL << area->shift to get size) nor does it sound like page 'order'.

Yeah good point.

Quoted 28 lines
Quoted 28 lines
mm/vmalloc.c +6 -3
 		}
 
 		kvfree(area->pages);
@@ -2541,14 +2590,17 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 				 pgprot_t prot, int node)
 {
 	struct page **pages;
+	unsigned long addr = (unsigned long)area->addr;
+	unsigned long size = get_vm_area_size(area);
+	unsigned int page_shift = area->page_shift;
+	unsigned int shift = page_shift + PAGE_SHIFT;
 	unsigned int nr_pages, array_size, i;
 	const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO;
 	const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN;
 	const gfp_t highmem_mask = (gfp_mask & (GFP_DMA | GFP_DMA32)) ?
-					0 :
-					__GFP_HIGHMEM;
+					0 : __GFP_HIGHMEM;
 
-	nr_pages = get_vm_area_size(area) >> PAGE_SHIFT;
+	nr_pages = size >> shift;
 	array_size = (nr_pages * sizeof(struct page *));
 
 	area->nr_pages = nr_pages;
@@ -2569,10 +2621,8 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 	for (i = 0; i < area->nr_pages; i++) {
 		struct page *page;
 
Quoted 8 lines

mm/vmalloc.c static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, · line 2625

Quoted 6 lines
mm/vmalloc.c +2 -4
-		if (node == NUMA_NO_NODE)
-			page = alloc_page(alloc_mask|highmem_mask);
-		else
-			page = alloc_pages_node(node, alloc_mask|highmem_mask, 0);
+		page = alloc_pages_node(node,
+				alloc_mask|highmem_mask, page_shift);

alloc_mask remains the exact same like before even for these high order pages.

Is there a problem there? I don't see.

Quoted 40 lines
Quoted 27 lines
mm/vmalloc.c +5 -2
 		if (unlikely(!page)) {
 			/* Successfully allocated i pages, free them in __vunmap() */
@@ -2584,8 +2634,9 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 			cond_resched();
 	}
 
-	if (map_vm_area(area, prot, pages))
+	if (vmap_hpages_range(addr, addr + size, prot, pages, page_shift) < 0)
 		goto fail;
+
 	return area->addr;
 
 fail:
@@ -2619,22 +2670,39 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
 			pgprot_t prot, unsigned long vm_flags, int node,
 			const void *caller)
 {
-	struct vm_struct *area;
+	struct vm_struct *area = NULL;
 	void *addr;
 	unsigned long real_size = size;
+	unsigned long real_align = align;
+	unsigned int shift = PAGE_SHIFT;
 
 	size = PAGE_ALIGN(size);
 	if (!size || (size >> PAGE_SHIFT) > totalram_pages())

mm/vmalloc.c void *__vmalloc_node_range(unsigned long size, unsigned long align, · line 2690

Quoted 10 lines
mm/vmalloc.c +8 -0
 		goto fail;
 
+	if (IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP)) {
+		unsigned long size_per_node;
+
+		size_per_node = size;
+		if (node == NUMA_NO_NODE)
+			size_per_node /= num_online_nodes();
+		if (size_per_node >= PMD_SIZE)
+			shift = PMD_SHIFT;

There are two problems here.

Quoted 2 lines

1. Should not size_per_node be aligned with PMD_SIZE to avoid wasting memory later because of alignment upwards (making it worse for NUMA_NO_NODE)

I'm not sure what you mean, it's just a heuristic to check for node interleaving, and use small pages if large can not interleave well.

Quoted 1 line

2. What about PUD_SIZE which is not considered here at all

Yeah, not doing PUD pages at all. It would be pretty trivial to add after PMD is working, but would it actually get used anywhere?

Quoted 10 lines

3. We should have similar knobs like ioremap controlling different size huge mappings

static int __read_mostly ioremap_p4d_capable;
static int __read_mostly ioremap_pud_capable;
static int __read_mostly ioremap_pmd_capable;
static int __read_mostly ioremap_huge_disabled;

while also giving arch a chance to weigh in through similar overrides like these.

arch_ioremap_[pud|pmd]_supported() ---> probably unifying it for vmalloc()

I'm not sure if that makes sense. IO mappings I could see maybe having some quirks or bugs or support issues. Cacheable mappings should be the "base" for supporting larger pages, if the platform has trouble with those then it should just avoid the feature I think.

Or is there a good reason to add the option? I don't mind, I just want to avoid proliferation.

Quoted 22 lines
Quoted 22 lines
mm/vmalloc.c +8 -1
+	}
+again:
+	align = max(real_align, 1UL << shift);
+	size = ALIGN(real_size, align);
+
 	area = __get_vm_area_node(size, align, VM_ALLOC | VM_UNINITIALIZED |
 				vm_flags, start, end, node, gfp_mask, caller);
 	if (!area)
 		goto fail;
 
+	area->page_shift = shift - PAGE_SHIFT;
+
 	addr = __vmalloc_area_node(area, gfp_mask, prot, node);
 	if (!addr)
-		return NULL;
+		goto fail;
 
 	/*
 	 * In this function, newly allocated vm_struct has VM_UNINITIALIZED
@@ -2648,8 +2716,16 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
 	return addr;
 
Quoted 10 lines

mm/vmalloc.c void *__vmalloc_node_range(unsigned long size, unsigned long align, · line 2722

Quoted 6 lines
mm/vmalloc.c +4 -1
 fail:
-	warn_alloc(gfp_mask, NULL,
+	if (shift == PMD_SHIFT) {
+		shift = PAGE_SHIFT;
+		goto again;
+	}

PUD_SHIFT should be accommodated here as well while falling back to lower mapping sizes in case previous allocation attempt fails.

This also would be part of PUD support.

Thanks,
Nick

#10 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Quoted 52 lines · Nicholas Piggin

Anshuman Khandual's on June 10, 2019 6:53 pm:

Quoted 42 lines · Nicholas Piggin
Quoted 2 lines

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them.

IIUC that extends HAVE_ARCH_HUGE_VMAP from iormap to vmalloc.

Quoted 5 lines

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

Sure will try this on arm64.

Quoted 28 lines
---
 include/asm-generic/4level-fixup.h |   1 +
 include/asm-generic/5level-fixup.h |   1 +
 include/linux/vmalloc.h            |   1 +
 mm/vmalloc.c                       | 132 +++++++++++++++++++++++------
 4 files changed, 107 insertions(+), 28 deletions(-)
include/asm-generic/4level-fixup.h +1 -0
index e3667c9a33a5..3cc65a4dd093 100644
@@ -20,6 +20,7 @@
 #define pud_none(pud)			0
 #define pud_bad(pud)			0
 #define pud_present(pud)		1
+#define pud_large(pud)			0
 #define pud_ERROR(pud)			do { } while (0)
 #define pud_clear(pud)			pgd_clear(pud)
 #define pud_val(pud)			pgd_val(pud)
include/asm-generic/5level-fixup.h +0 -0
index bb6cb347018c..c4377db09a4f 100644
@@ -22,6 +22,7 @@
 #define p4d_none(p4d)			0
 #define p4d_bad(p4d)			0
Quoted 8 lines

include/asm-generic/5level-fixup.h line 28

Quoted 5 lines
include/asm-generic/5level-fixup.h +1 -0
 #define p4d_present(p4d)		1
+#define p4d_large(p4d)			0
 #define p4d_ERROR(p4d)			do { } while (0)
 #define p4d_clear(p4d)			pgd_clear(p4d)
 #define p4d_val(p4d)			pgd_val(p4d)

Both of these are required from vmalloc_to_page() which as per a later comment should be part of a prerequisite patch before this series.

Quoted 1 line

I'm not sure what you mean. This patch is where they get used.

In case you move out vmalloc_to_page() changes to a separate patch.

Quoted 18 lines

Possibly I could split this and the vmalloc_to_page change out. I'll consider it.

Quoted 5 lines
Quoted 5 lines
include/linux/vmalloc.h +0 -0
index 812bea5866d6..4c92dc608928 100644
@@ -42,6 +42,7 @@ struct vm_struct {
Quoted 8 lines

include/linux/vmalloc.h struct vm_struct { · line 45

Quoted 4 lines
include/linux/vmalloc.h +1 -0
 	unsigned long		size;
 	unsigned long		flags;
 	struct page		**pages;
+	unsigned int		page_shift;

So the entire vm_struct will be mapped with a single page_shift. It cannot have mix and match mappings with PAGE_SIZE, PMD_SIZE, PUD_SIZE etc in case the allocation fails for larger ones, falling back etc what over other reasons.

Quoted 4 lines

For now, yes. I have a bit of follow up work to improve that and make it able to fall back, but it's a bit more churn and not a significant benefit just yet because there are not a lot of very large vmallocs (except the early hashes which can be satisfied with large allocs).

Right but it will make this new feature complete like ioremap which logically supports till P4D (though AFAICT not used). If there are no actual vmalloc requests that large it is fine. Allocation attempts will start from the page table level depending on the requested size. It is better to have PUD/P4D considerations now rather than trying to after fit it later.

Quoted 21 lines
Quoted 21 lines
Quoted 21 lines
include/linux/vmalloc.h +0 -0
 	unsigned int		nr_pages;
 	phys_addr_t		phys_addr;
 	const void		*caller;
mm/vmalloc.c +1 -0
index dd27cfb29b10..0cf8e861caeb 100644
@@ -36,6 +36,7 @@
 #include <linux/rbtree_augmented.h>
 
 #include <linux/uaccess.h>
+#include <asm/pgtable.h>
 #include <asm/tlbflush.h>
 #include <asm/shmparam.h>
 
@@ -440,6 +441,41 @@ static int vmap_pages_range(unsigned long start, unsigned long end,
 	return ret;
 }
 
Quoted 7 lines
Quoted 4 lines

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 445

Quoted 2 lines
mm/vmalloc.c +2 -0
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+static int vmap_hpages_range(unsigned long start, unsigned long end,

A small nit (if you agree) s/hpages/huge_pages/

Hmm. It's not actually a good function name because it can do small pages as well. vmap_pages_size_range or something may be better.

Right.

Quoted 15 lines
Quoted 5 lines

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 446

Quoted 3 lines
mm/vmalloc.c +1 -0
+				   pgprot_t prot, struct page **pages,

Re-order (prot <---> pages) just to follow the standard like before.

Will do.

Quoted 6 lines

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 450

Quoted 4 lines
mm/vmalloc.c +4 -0
+				   unsigned int page_shift)
+{
+	unsigned long addr = start;
+	unsigned int i, nr = (end - start) >> (PAGE_SHIFT + page_shift);

s/nr/nr_huge_pages ?

Quoted 7 lines

Sure.

Quoted 2 lines

Also should not we check for the alignment of the range [start...end] with respect to (1UL << [PAGE_SHIFT + page_shift]).

The caller should if it specifies large page. Could check and -EINVAL for incorrect alignment.

That might be a good check here.

Quoted 52 lines
Quoted 52 lines
Quoted 42 lines
mm/vmalloc.c +29 -1
+
+	for (i = 0; i < nr; i++) {
+		int err;
+
+		err = vmap_range_noflush(addr,
+					addr + (PAGE_SIZE << page_shift),
+					__pa(page_address(pages[i])), prot,
+					PAGE_SHIFT + page_shift);
+		if (err)
+			return err;
+
+		addr += PAGE_SIZE << page_shift;
+	}
+	flush_cache_vmap(start, end);
+
+	return nr;
+}
+#else
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+			   pgprot_t prot, struct page **pages,
+			   unsigned int page_shift)
+{
+	BUG_ON(page_shift != PAGE_SIZE);
+	return vmap_pages_range(start, end, prot, pages);
+}
+#endif
+
+
 int is_vmalloc_or_module_addr(const void *x)
 {
 	/*
@@ -462,7 +498,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 {
 	unsigned long addr = (unsigned long) vmalloc_addr;
 	struct page *page = NULL;
-	pgd_t *pgd = pgd_offset_k(addr);
+	pgd_t *pgd;
 	p4d_t *p4d;
 	pud_t *pud;
 	pmd_t *pmd;

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 516

Quoted 7 lines
mm/vmalloc.c +2 -0
@@ -474,27 +510,38 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	 */
 	VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr));
 
+	pgd = pgd_offset_k(addr);
 	if (pgd_none(*pgd))
 		return NULL;
+

Small nit. Stray line here.

Quoted 4 lines
Quoted 1 line

'pgd' related changes here seem to be just cleanups and should not part of this patch.

Yeah I figure it doesn't matter to make small changes close by, but maybe that's more frowned upon now for git blame?

Right. But I guess it should be okay if you can make vmalloc_to_page() changes as a separate patch. This patch which adds a new feature should not have any clean ups IMHO.

Quoted 24 lines
Quoted 12 lines
Quoted 12 lines
mm/vmalloc.c +6 -1
 	p4d = p4d_offset(pgd, addr);
 	if (p4d_none(*p4d))
 		return NULL;
-	pud = pud_offset(p4d, addr);
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (p4d_large(*p4d))
+		return p4d_page(*p4d) + ((addr & ~P4D_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(p4d_bad(*p4d)))
+		return NULL;
 
Quoted 11 lines

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 491

Quoted 8 lines
mm/vmalloc.c +0 -8
-	/*
-	 * Don't dereference bad PUD or PMD (below) entries. This will also
-	 * identify huge mappings, which we may encounter on architectures
-	 * that define CONFIG_HAVE_ARCH_HUGE_VMAP=y. Such regions will be
-	 * identified as vmalloc addresses by is_vmalloc_addr(), but are
-	 * not [unambiguously] associated with a struct page, so there is
-	 * no correct value to return for them.
-	 */

What changed the situation so that we could return struct page for a huge mapping now ?

Quoted 2 lines

For the PUD case? Nothing changed, per se, we I just calculate the correct struct page now, so I may return it.

I was just curious what prevented this earlier (before this series). The comment here and commit message which added this change making me wonder what was the reason for not doing this earlier.

Quoted 10 lines
Quoted 6 lines

AFAICT even after this patch, PUD/P4D level huge pages can only be created with ioremap_page_range() not with vmalloc() which creates PMD sized mappings only. Hence if it's okay to dereference struct page of a huge mapping (not withstanding the comment here) it should be part of an earlier patch fixing it first for existing ioremap_page_range() huge mappings.

Possibly yes, we can consider 029c54b095995 to be a band-aid for huge vmaps which is fixed properly by this change, in which case it could make sense to break this into its own patch.

On arm64 [pud|pmd]_bad() calls out huge mappings at PUD or PMD. I still wonder what Ard (copied him now) meant by "not [unambiguously] associated with a struct page". He also mentioned about compound pages in the commit message. Anyways these makes sense (fetching the struct page) unless I am missing something. But should be part of a separate patch.

pXd_page(*pXd) + ((addr & ~PXD_MASK) >> PAGE_SHIFT)

Quoted 36 lines
Quoted 36 lines
Quoted 15 lines
mm/vmalloc.c +9 -2
-	WARN_ON_ONCE(pud_bad(*pud));
-	if (pud_none(*pud) || pud_bad(*pud))
+	pud = pud_offset(p4d, addr);
+	if (pud_none(*pud))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pud_large(*pud))
+		return pud_page(*pud) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pud_bad(*pud)))
 		return NULL;
+
 	pmd = pmd_offset(pud, addr);

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 545

Quoted 10 lines
mm/vmalloc.c +7 -2
-	WARN_ON_ONCE(pmd_bad(*pmd));
-	if (pmd_none(*pmd) || pmd_bad(*pmd))
+	if (pmd_none(*pmd))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pmd_large(*pmd))
+		return pmd_page(*pmd) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pmd_bad(*pmd)))
 		return NULL;

At each page table level, we are checking in this order

pXX_none() --> pXX_large() --> pXX_bad()

Are not these alternative orders bit better

pXX_bad() --> pXX_none() --> pXX_large()

Or

Quoted 7 lines
Quoted 3 lines

pXX_none() --> pXX_bad() --> pXX_large()

Checking for pXX_bad() at the end does not make much sense.

Yeah the order tends to go none->bad. It's not 100% clear we can test bad before none (at least it goes against convention). But good point should be changed to your last sequence I think.

Sure.

Quoted 81 lines
Quoted 11 lines

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 552

Quoted 9 lines
mm/vmalloc.c +1 -0
 	ptep = pte_offset_map(pmd, addr);
@@ -502,6 +549,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	if (pte_present(pte))
 		page = pte_page(pte);
 	pte_unmap(ptep);
+

Small nit. Stray line here.

I don't mind adding some lines here and there, like here. It is an unrelated (alleged-)cleanup though.

Quoted 19 lines
Quoted 19 lines
mm/vmalloc.c +3 -2
 	return page;
 }
 EXPORT_SYMBOL(vmalloc_to_page);
@@ -2185,8 +2233,9 @@ static struct vm_struct *__get_vm_area_node(unsigned long size,
 		return NULL;
 
 	if (flags & VM_IOREMAP)
-		align = 1ul << clamp_t(int, get_count_order_long(size),
-				       PAGE_SHIFT, IOREMAP_MAX_ORDER);
+		align = max(align,
+				1ul << clamp_t(int, get_count_order_long(size),
+				       PAGE_SHIFT, IOREMAP_MAX_ORDER));
 
 	area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node);
 	if (unlikely(!area))
@@ -2398,7 +2447,7 @@ static void __vunmap(const void *addr, int deallocate_pages)
 			struct page *page = area->pages[i];
 
Quoted 7 lines

mm/vmalloc.c static void __vunmap(const void *addr, int deallocate_pages) · line 2450

Quoted 3 lines
mm/vmalloc.c +1 -1
 			BUG_ON(!page);
-			__free_pages(page, 0);
+			__free_pages(page, area->page_shift);

area->page_shift' turns out to be effective page order. I think the name here is bit misleading. s/page_shift/page_order or nr_pages should be better IMHO. page_shift is not actual shift (1UL << area->shift to get size) nor does it sound like page 'order'.

Yeah good point.

Quoted 28 lines
Quoted 28 lines
mm/vmalloc.c +6 -3
 		}
 
 		kvfree(area->pages);
@@ -2541,14 +2590,17 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 				 pgprot_t prot, int node)
 {
 	struct page **pages;
+	unsigned long addr = (unsigned long)area->addr;
+	unsigned long size = get_vm_area_size(area);
+	unsigned int page_shift = area->page_shift;
+	unsigned int shift = page_shift + PAGE_SHIFT;
 	unsigned int nr_pages, array_size, i;
 	const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO;
 	const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN;
 	const gfp_t highmem_mask = (gfp_mask & (GFP_DMA | GFP_DMA32)) ?
-					0 :
-					__GFP_HIGHMEM;
+					0 : __GFP_HIGHMEM;
 
-	nr_pages = get_vm_area_size(area) >> PAGE_SHIFT;
+	nr_pages = size >> shift;
 	array_size = (nr_pages * sizeof(struct page *));
 
 	area->nr_pages = nr_pages;
@@ -2569,10 +2621,8 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 	for (i = 0; i < area->nr_pages; i++) {
 		struct page *page;
 
Quoted 8 lines

mm/vmalloc.c static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, · line 2625

Quoted 6 lines
mm/vmalloc.c +2 -4
-		if (node == NUMA_NO_NODE)
-			page = alloc_page(alloc_mask|highmem_mask);
-		else
-			page = alloc_pages_node(node, alloc_mask|highmem_mask, 0);
+		page = alloc_pages_node(node,
+				alloc_mask|highmem_mask, page_shift);

alloc_mask remains the exact same like before even for these high order pages.

Quoted 1 line

Is there a problem there? I don't see.

There is no problem. I justed noted it.

Quoted 44 lines
Quoted 41 lines
Quoted 28 lines
mm/vmalloc.c +5 -2
 		if (unlikely(!page)) {
 			/* Successfully allocated i pages, free them in __vunmap() */
@@ -2584,8 +2634,9 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 			cond_resched();
 	}
 
-	if (map_vm_area(area, prot, pages))
+	if (vmap_hpages_range(addr, addr + size, prot, pages, page_shift) < 0)
 		goto fail;
+
 	return area->addr;
 
 fail:
@@ -2619,22 +2670,39 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
 			pgprot_t prot, unsigned long vm_flags, int node,
 			const void *caller)
 {
-	struct vm_struct *area;
+	struct vm_struct *area = NULL;
 	void *addr;
 	unsigned long real_size = size;
+	unsigned long real_align = align;
+	unsigned int shift = PAGE_SHIFT;
 
 	size = PAGE_ALIGN(size);
 	if (!size || (size >> PAGE_SHIFT) > totalram_pages())

mm/vmalloc.c void *__vmalloc_node_range(unsigned long size, unsigned long align, · line 2690

Quoted 10 lines
mm/vmalloc.c +8 -0
 		goto fail;
 
+	if (IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP)) {
+		unsigned long size_per_node;
+
+		size_per_node = size;
+		if (node == NUMA_NO_NODE)
+			size_per_node /= num_online_nodes();
+		if (size_per_node >= PMD_SIZE)
+			shift = PMD_SHIFT;

There are two problems here.

Quoted 2 lines

1. Should not size_per_node be aligned with PMD_SIZE to avoid wasting memory later because of alignment upwards (making it worse for NUMA_NO_NODE)

Quoted 7 lines

I'm not sure what you mean, it's just a heuristic to check for node interleaving, and use small pages if large can not interleave well.

Quoted 1 line

2. What about PUD_SIZE which is not considered here at all

Yeah, not doing PUD pages at all. It would be pretty trivial to add after PMD is working, but would it actually get used anywhere?

But it should make this feature logically complete. Allocation attempts can start at right pgtable level depending on the requested size. I dont think it will have any performance impact or something.

Quoted 12 lines
Quoted 11 lines

3. We should have similar knobs like ioremap controlling different size huge mappings

static int __read_mostly ioremap_p4d_capable;
static int __read_mostly ioremap_pud_capable;
static int __read_mostly ioremap_pmd_capable;
static int __read_mostly ioremap_huge_disabled;

while also giving arch a chance to weigh in through similar overrides like these.

arch_ioremap_[pud|pmd]_supported() ---> probably unifying it for vmalloc()

Quoted 7 lines

I'm not sure if that makes sense. IO mappings I could see maybe having some quirks or bugs or support issues. Cacheable mappings should be the "base" for supporting larger pages, if the platform has trouble with those then it should just avoid the feature I think.

Or is there a good reason to add the option? I don't mind, I just want to avoid proliferation.

We might need atleast for the potential problem on arm64 as discussed on the other thread.

author reviewer
#11 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Anshuman Khandual's on June 11, 2019 4:59 pm:

Quoted 53 lines · Nicholas Piggin
Quoted 52 lines

Anshuman Khandual's on June 10, 2019 6:53 pm:

Quoted 50 lines · Nicholas Piggin
Quoted 2 lines

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them.

IIUC that extends HAVE_ARCH_HUGE_VMAP from iormap to vmalloc.

Quoted 5 lines

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

Sure will try this on arm64.

Quoted 26 lines
---
 include/asm-generic/4level-fixup.h |   1 +
 include/asm-generic/5level-fixup.h |   1 +
 include/linux/vmalloc.h            |   1 +
 mm/vmalloc.c                       | 132 +++++++++++++++++++++++------
 4 files changed, 107 insertions(+), 28 deletions(-)
include/asm-generic/4level-fixup.h +1 -0
index e3667c9a33a5..3cc65a4dd093 100644
@@ -20,6 +20,7 @@
 #define pud_none(pud)			0
 #define pud_bad(pud)			0
 #define pud_present(pud)		1
+#define pud_large(pud)			0
 #define pud_ERROR(pud)			do { } while (0)
 #define pud_clear(pud)			pgd_clear(pud)
 #define pud_val(pud)			pgd_val(pud)
include/asm-generic/5level-fixup.h +0 -0
index bb6cb347018c..c4377db09a4f 100644

include/asm-generic/5level-fixup.h line 28

Quoted 7 lines
include/asm-generic/5level-fixup.h +1 -0
@@ -22,6 +22,7 @@
 #define p4d_none(p4d)			0
 #define p4d_bad(p4d)			0
 #define p4d_present(p4d)		1
+#define p4d_large(p4d)			0
 #define p4d_ERROR(p4d)			do { } while (0)
 #define p4d_clear(p4d)			pgd_clear(p4d)
 #define p4d_val(p4d)			pgd_val(p4d)

Both of these are required from vmalloc_to_page() which as per a later comment should be part of a prerequisite patch before this series.

Quoted 3 lines
Quoted 1 line

I'm not sure what you mean. This patch is where they get used.

In case you move out vmalloc_to_page() changes to a separate patch.

Sorry for the delay in reply.

I'll split this and see if we might be able to get it into next merge window. I can have another try at the huge vmalloc patch after that.

Quoted 24 lines
Quoted 19 lines

Possibly I could split this and the vmalloc_to_page change out. I'll consider it.

Quoted 5 lines
Quoted 5 lines
include/linux/vmalloc.h +0 -0
index 812bea5866d6..4c92dc608928 100644
@@ -42,6 +42,7 @@ struct vm_struct {
Quoted 8 lines

include/linux/vmalloc.h struct vm_struct { · line 45

Quoted 4 lines
include/linux/vmalloc.h +1 -0
 	unsigned long		size;
 	unsigned long		flags;
 	struct page		**pages;
+	unsigned int		page_shift;

So the entire vm_struct will be mapped with a single page_shift. It cannot have mix and match mappings with PAGE_SIZE, PMD_SIZE, PUD_SIZE etc in case the allocation fails for larger ones, falling back etc what over other reasons.

Quoted 4 lines

For now, yes. I have a bit of follow up work to improve that and make it able to fall back, but it's a bit more churn and not a significant benefit just yet because there are not a lot of very large vmallocs (except the early hashes which can be satisfied with large allocs).

Quoted 5 lines

Right but it will make this new feature complete like ioremap which logically supports till P4D (though AFAICT not used). If there are no actual vmalloc requests that large it is fine. Allocation attempts will start from the page table level depending on the requested size. It is better to have PUD/P4D considerations now rather than trying to after fit it later.

I've considered them, which is why e.g., a shift gets passed around rather than a bool for small/large.

I won't over complicate this page array data structure for something that may never be supported though. I think we may actually be better moving away from it in the vmalloc code and just referencing pages from the page tables, so it's just something we can cross when we get to it.

Quoted 7 lines
Quoted 5 lines
Quoted 2 lines

Also should not we check for the alignment of the range [start...end] with respect to (1UL << [PAGE_SHIFT + page_shift]).

The caller should if it specifies large page. Could check and -EINVAL for incorrect alignment.

That might be a good check here.

Will add.

Quoted 11 lines
Quoted 11 lines
Quoted 11 lines

include/linux/vmalloc.h struct page *vmalloc_to_page(const void *vmalloc_addr) · line 516

Quoted 8 lines
include/linux/vmalloc.h +2 -0
@@ -474,27 +510,38 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	 */
 	VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr));
 
+	pgd = pgd_offset_k(addr);
 	if (pgd_none(*pgd))
 		return NULL;
+

Small nit. Stray line here.

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

'pgd' related changes here seem to be just cleanups and should not part of this patch.

Yeah I figure it doesn't matter to make small changes close by, but maybe that's more frowned upon now for git blame?

Right. But I guess it should be okay if you can make vmalloc_to_page() changes as a separate patch. This patch which adds a new feature should not have any clean ups IMHO.

Well... that alone would be a new feature too. Or could be considered a bug fix, which makes it even more important not to contain superfluous changes.

Is there a real prohibition on small slightly peripheral tidying like this? I don't think I'd bother sending a lone patch just to change a couple lines of spacing.

Quoted 23 lines
Quoted 23 lines
Quoted 22 lines
Quoted 9 lines
include/linux/vmalloc.h +5 -1
 	p4d = p4d_offset(pgd, addr);
 	if (p4d_none(*p4d))
 		return NULL;
-	pud = pud_offset(p4d, addr);
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (p4d_large(*p4d))
+		return p4d_page(*p4d) + ((addr & ~P4D_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(p4d_bad(*p4d)))

include/linux/vmalloc.h struct page *vmalloc_to_page(const void *vmalloc_addr) · line 491

Quoted 10 lines
include/linux/vmalloc.h +1 -8
+		return NULL;
 
-	/*
-	 * Don't dereference bad PUD or PMD (below) entries. This will also
-	 * identify huge mappings, which we may encounter on architectures
-	 * that define CONFIG_HAVE_ARCH_HUGE_VMAP=y. Such regions will be
-	 * identified as vmalloc addresses by is_vmalloc_addr(), but are
-	 * not [unambiguously] associated with a struct page, so there is
-	 * no correct value to return for them.
-	 */

What changed the situation so that we could return struct page for a huge mapping now ?

Quoted 6 lines
Quoted 2 lines

For the PUD case? Nothing changed, per se, we I just calculate the correct struct page now, so I may return it.

I was just curious what prevented this earlier (before this series). The comment here and commit message which added this change making me wonder what was the reason for not doing this earlier.

Just not implemented I guess.

Quoted 10 lines
Quoted 9 lines
Quoted 5 lines

AFAICT even after this patch, PUD/P4D level huge pages can only be created with ioremap_page_range() not with vmalloc() which creates PMD sized mappings only. Hence if it's okay to dereference struct page of a huge mapping (not withstanding the comment here) it should be part of an earlier patch fixing it first for existing ioremap_page_range() huge mappings.

Possibly yes, we can consider 029c54b095995 to be a band-aid for huge vmaps which is fixed properly by this change, in which case it could make sense to break this into its own patch.

Quoted 5 lines

On arm64 [pud|pmd]_bad() calls out huge mappings at PUD or PMD. I still wonder what Ard (copied him now) meant by "not [unambiguously] associated with a struct page". He also mentioned about compound pages in the commit message. Anyways these makes sense (fetching the struct page) unless I am missing something. But should be part of a separate patch.

I do somewhat see the intention of the commit message, but if we consider the vmap/iomap layer's choice of page size as transparent to the caller, and the vmalloc_to_page API has always been specifically interested in the PAGE_SIZE struct page, then my patch is fine and introduces no problems. It restores the API functionality to be the same regardless of whether small or large pages were used for the actual mapping.

Quoted 17 lines
Quoted 17 lines
Quoted 11 lines

include/linux/vmalloc.h struct page *vmalloc_to_page(const void *vmalloc_addr) · line 534

Quoted 8 lines
include/linux/vmalloc.h +8 -0
+	if (IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP)) {
+		unsigned long size_per_node;
+
+		size_per_node = size;
+		if (node == NUMA_NO_NODE)
+			size_per_node /= num_online_nodes();
+		if (size_per_node >= PMD_SIZE)
+			shift = PMD_SHIFT;

There are two problems here.

Quoted 2 lines

1. Should not size_per_node be aligned with PMD_SIZE to avoid wasting memory later because of alignment upwards (making it worse for NUMA_NO_NODE)

I'm not sure what you mean, it's just a heuristic to check for node interleaving, and use small pages if large can not interleave well.

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

2. What about PUD_SIZE which is not considered here at all

Yeah, not doing PUD pages at all. It would be pretty trivial to add after PMD is working, but would it actually get used anywhere?

But it should make this feature logically complete. Allocation attempts can start at right pgtable level depending on the requested size. I dont think it will have any performance impact or something.

I disagree that's necessary or desirable for PMD support here. Sure an arch might have PUD size within MAX_ORDER and implement that, but it's just something that can be implemented when the time comes.

There's nothing about this patch that hinders being extendedto PUD level I just won't add code that's not used and I can't test.

Thanks for the detailed review, I appreciate it.

Thanks,
Nick

reviewer
#12 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

In reply to Nicholas Piggin

Hi,

Quoted 7 lines · Nicholas Piggin

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

Do you happen to know which vmalloc mappings these get used for in the above case? Where do we see vmalloc mappings that large?

I'm worried as to how this would interact with the set_memory_*() functions, as on arm64 those can only operate on page-granular mappings. Those may need fixing up to handle huge mappings; certainly if the above is all for modules.

Thanks,
Mark.

Quoted 302 lines
---
 include/asm-generic/4level-fixup.h |   1 +
 include/asm-generic/5level-fixup.h |   1 +
 include/linux/vmalloc.h            |   1 +
 mm/vmalloc.c                       | 132 +++++++++++++++++++++++------
 4 files changed, 107 insertions(+), 28 deletions(-)
include/asm-generic/4level-fixup.h +1 -0
index e3667c9a33a5..3cc65a4dd093 100644
@@ -20,6 +20,7 @@
 #define pud_none(pud)			0
 #define pud_bad(pud)			0
 #define pud_present(pud)		1
+#define pud_large(pud)			0
 #define pud_ERROR(pud)			do { } while (0)
 #define pud_clear(pud)			pgd_clear(pud)
 #define pud_val(pud)			pgd_val(pud)
include/asm-generic/5level-fixup.h +1 -0
index bb6cb347018c..c4377db09a4f 100644
@@ -22,6 +22,7 @@
 #define p4d_none(p4d)			0
 #define p4d_bad(p4d)			0
 #define p4d_present(p4d)		1
+#define p4d_large(p4d)			0
 #define p4d_ERROR(p4d)			do { } while (0)
 #define p4d_clear(p4d)			pgd_clear(p4d)
 #define p4d_val(p4d)			pgd_val(p4d)
include/linux/vmalloc.h +1 -0
index 812bea5866d6..4c92dc608928 100644
@@ -42,6 +42,7 @@ struct vm_struct {
 	unsigned long		size;
 	unsigned long		flags;
 	struct page		**pages;
+	unsigned int		page_shift;
 	unsigned int		nr_pages;
 	phys_addr_t		phys_addr;
 	const void		*caller;
mm/vmalloc.c +104 -28
index dd27cfb29b10..0cf8e861caeb 100644
@@ -36,6 +36,7 @@
 #include <linux/rbtree_augmented.h>
 
 #include <linux/uaccess.h>
+#include <asm/pgtable.h>
 #include <asm/tlbflush.h>
 #include <asm/shmparam.h>
 
@@ -440,6 +441,41 @@ static int vmap_pages_range(unsigned long start, unsigned long end,
 	return ret;
 }
 
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+				   pgprot_t prot, struct page **pages,
+				   unsigned int page_shift)
+{
+	unsigned long addr = start;
+	unsigned int i, nr = (end - start) >> (PAGE_SHIFT + page_shift);
+
+	for (i = 0; i < nr; i++) {
+		int err;
+
+		err = vmap_range_noflush(addr,
+					addr + (PAGE_SIZE << page_shift),
+					__pa(page_address(pages[i])), prot,
+					PAGE_SHIFT + page_shift);
+		if (err)
+			return err;
+
+		addr += PAGE_SIZE << page_shift;
+	}
+	flush_cache_vmap(start, end);
+
+	return nr;
+}
+#else
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+			   pgprot_t prot, struct page **pages,
+			   unsigned int page_shift)
+{
+	BUG_ON(page_shift != PAGE_SIZE);
+	return vmap_pages_range(start, end, prot, pages);
+}
+#endif
+
+
 int is_vmalloc_or_module_addr(const void *x)
 {
 	/*
@@ -462,7 +498,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 {
 	unsigned long addr = (unsigned long) vmalloc_addr;
 	struct page *page = NULL;
-	pgd_t *pgd = pgd_offset_k(addr);
+	pgd_t *pgd;
 	p4d_t *p4d;
 	pud_t *pud;
 	pmd_t *pmd;
@@ -474,27 +510,38 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	 */
 	VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr));
 
+	pgd = pgd_offset_k(addr);
 	if (pgd_none(*pgd))
 		return NULL;
+
 	p4d = p4d_offset(pgd, addr);
 	if (p4d_none(*p4d))
 		return NULL;
-	pud = pud_offset(p4d, addr);
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (p4d_large(*p4d))
+		return p4d_page(*p4d) + ((addr & ~P4D_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(p4d_bad(*p4d)))
+		return NULL;
 
-	/*
-	 * Don't dereference bad PUD or PMD (below) entries. This will also
-	 * identify huge mappings, which we may encounter on architectures
-	 * that define CONFIG_HAVE_ARCH_HUGE_VMAP=y. Such regions will be
-	 * identified as vmalloc addresses by is_vmalloc_addr(), but are
-	 * not [unambiguously] associated with a struct page, so there is
-	 * no correct value to return for them.
-	 */
-	WARN_ON_ONCE(pud_bad(*pud));
-	if (pud_none(*pud) || pud_bad(*pud))
+	pud = pud_offset(p4d, addr);
+	if (pud_none(*pud))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pud_large(*pud))
+		return pud_page(*pud) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pud_bad(*pud)))
 		return NULL;
+
 	pmd = pmd_offset(pud, addr);
-	WARN_ON_ONCE(pmd_bad(*pmd));
-	if (pmd_none(*pmd) || pmd_bad(*pmd))
+	if (pmd_none(*pmd))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pmd_large(*pmd))
+		return pmd_page(*pmd) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pmd_bad(*pmd)))
 		return NULL;
 
 	ptep = pte_offset_map(pmd, addr);
@@ -502,6 +549,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
 	if (pte_present(pte))
 		page = pte_page(pte);
 	pte_unmap(ptep);
+
 	return page;
 }
 EXPORT_SYMBOL(vmalloc_to_page);
@@ -2185,8 +2233,9 @@ static struct vm_struct *__get_vm_area_node(unsigned long size,
 		return NULL;
 
 	if (flags & VM_IOREMAP)
-		align = 1ul << clamp_t(int, get_count_order_long(size),
-				       PAGE_SHIFT, IOREMAP_MAX_ORDER);
+		align = max(align,
+				1ul << clamp_t(int, get_count_order_long(size),
+				       PAGE_SHIFT, IOREMAP_MAX_ORDER));
 
 	area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node);
 	if (unlikely(!area))
@@ -2398,7 +2447,7 @@ static void __vunmap(const void *addr, int deallocate_pages)
 			struct page *page = area->pages[i];
 
 			BUG_ON(!page);
-			__free_pages(page, 0);
+			__free_pages(page, area->page_shift);
 		}
 
 		kvfree(area->pages);
@@ -2541,14 +2590,17 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 				 pgprot_t prot, int node)
 {
 	struct page **pages;
+	unsigned long addr = (unsigned long)area->addr;
+	unsigned long size = get_vm_area_size(area);
+	unsigned int page_shift = area->page_shift;
+	unsigned int shift = page_shift + PAGE_SHIFT;
 	unsigned int nr_pages, array_size, i;
 	const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO;
 	const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN;
 	const gfp_t highmem_mask = (gfp_mask & (GFP_DMA | GFP_DMA32)) ?
-					0 :
-					__GFP_HIGHMEM;
+					0 : __GFP_HIGHMEM;
 
-	nr_pages = get_vm_area_size(area) >> PAGE_SHIFT;
+	nr_pages = size >> shift;
 	array_size = (nr_pages * sizeof(struct page *));
 
 	area->nr_pages = nr_pages;
@@ -2569,10 +2621,8 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 	for (i = 0; i < area->nr_pages; i++) {
 		struct page *page;
 
-		if (node == NUMA_NO_NODE)
-			page = alloc_page(alloc_mask|highmem_mask);
-		else
-			page = alloc_pages_node(node, alloc_mask|highmem_mask, 0);
+		page = alloc_pages_node(node,
+				alloc_mask|highmem_mask, page_shift);
 
 		if (unlikely(!page)) {
 			/* Successfully allocated i pages, free them in __vunmap() */
@@ -2584,8 +2634,9 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
 			cond_resched();
 	}
 
-	if (map_vm_area(area, prot, pages))
+	if (vmap_hpages_range(addr, addr + size, prot, pages, page_shift) < 0)
 		goto fail;
+
 	return area->addr;
 
 fail:
@@ -2619,22 +2670,39 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
 			pgprot_t prot, unsigned long vm_flags, int node,
 			const void *caller)
 {
-	struct vm_struct *area;
+	struct vm_struct *area = NULL;
 	void *addr;
 	unsigned long real_size = size;
+	unsigned long real_align = align;
+	unsigned int shift = PAGE_SHIFT;
 
 	size = PAGE_ALIGN(size);
 	if (!size || (size >> PAGE_SHIFT) > totalram_pages())
 		goto fail;
 
+	if (IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP)) {
+		unsigned long size_per_node;
+
+		size_per_node = size;
+		if (node == NUMA_NO_NODE)
+			size_per_node /= num_online_nodes();
+		if (size_per_node >= PMD_SIZE)
+			shift = PMD_SHIFT;
+	}
+again:
+	align = max(real_align, 1UL << shift);
+	size = ALIGN(real_size, align);
+
 	area = __get_vm_area_node(size, align, VM_ALLOC | VM_UNINITIALIZED |
 				vm_flags, start, end, node, gfp_mask, caller);
 	if (!area)
 		goto fail;
 
+	area->page_shift = shift - PAGE_SHIFT;
+
 	addr = __vmalloc_area_node(area, gfp_mask, prot, node);
 	if (!addr)
-		return NULL;
+		goto fail;
 
 	/*
 	 * In this function, newly allocated vm_struct has VM_UNINITIALIZED
@@ -2648,8 +2716,16 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
 	return addr;
 
 fail:
-	warn_alloc(gfp_mask, NULL,
+	if (shift == PMD_SHIFT) {
+		shift = PAGE_SHIFT;
+		goto again;
+	}
+
+	if (!area) {
+		/* Warn for area allocation, page allocations already warn */
+		warn_alloc(gfp_mask, NULL,
 			  "vmalloc: allocation failure: %lu bytes", real_size);
+	}
 	return NULL;
 }
 
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author reviewer
#13 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Mark Rutland's on June 11, 2019 12:10 am:

Quoted 6 lines

Hi,

Quoted 3 lines · Nicholas Piggin

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them

Quoted 7 lines
Quoted 4 lines

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

Do you happen to know which vmalloc mappings these get used for in the above case? Where do we see vmalloc mappings that large?

Large module vmalloc could be subject to huge mappings.

Quoted 4 lines

I'm worried as to how this would interact with the set_memory_*() functions, as on arm64 those can only operate on page-granular mappings. Those may need fixing up to handle huge mappings; certainly if the above is all for modules.

Good point, that looks like it would break on arm64 at least. I'll work on it. We may have to make this opt in beyond HUGE_VMAP.

Thanks,
Nick

#14 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Quoted 17 lines · Nicholas Piggin

Mark Rutland's on June 11, 2019 12:10 am:

Quoted 6 lines

Hi,

Quoted 3 lines · Nicholas Piggin

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them

Quoted 7 lines
Quoted 4 lines

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

Do you happen to know which vmalloc mappings these get used for in the above case? Where do we see vmalloc mappings that large?

Large module vmalloc could be subject to huge mappings.

Quoted 7 lines
Quoted 4 lines

I'm worried as to how this would interact with the set_memory_*() functions, as on arm64 those can only operate on page-granular mappings. Those may need fixing up to handle huge mappings; certainly if the above is all for modules.

Good point, that looks like it would break on arm64 at least. I'll work on it. We may have to make this opt in beyond HUGE_VMAP.

This is another reason we might need to have an arch opt-ins like the one I mentioned before.

author reviewer
#15 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Anshuman Khandual's on June 11, 2019 4:17 pm:

Quoted 25 lines · Nicholas Piggin
Quoted 22 lines

Mark Rutland's on June 11, 2019 12:10 am:

Quoted 6 lines

Hi,

Quoted 3 lines · Nicholas Piggin

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them

Quoted 7 lines
Quoted 4 lines

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

Do you happen to know which vmalloc mappings these get used for in the above case? Where do we see vmalloc mappings that large?

Large module vmalloc could be subject to huge mappings.

Quoted 4 lines

I'm worried as to how this would interact with the set_memory_*() functions, as on arm64 those can only operate on page-granular mappings. Those may need fixing up to handle huge mappings; certainly if the above is all for modules.

Quoted 6 lines
Quoted 2 lines

Good point, that looks like it would break on arm64 at least. I'll work on it. We may have to make this opt in beyond HUGE_VMAP.

This is another reason we might need to have an arch opt-ins like the one I mentioned before.

Let's try to get the precursor stuff like page table functions and vmalloc_to_page in this merge window, and then concentrate on the huge vmalloc support issues after that.

Christophe points out that powerpc is likely to have a similar problem which I didn't realise, so I'll re think it.

Thanks,
Nick

#16 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

In reply to Nicholas Piggin

Le 10/06/2019 à 06:38, Nicholas Piggin a écrit :

Quoted 2 lines

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them

Will this be compatible with Russell's series https://patchwork.ozlabs.org/patch/1099857/ for the implementation of STRICT_MODULE_RWX ? I see that apply_to_page_range() have things like BUG_ON(pud_huge(*pud));

Might also be an issue for arm64 as I think Russell's implementation comes from there.

Quoted 86 lines

This brings dTLB misses for linux kernel tree git diff from 45,000 to 8,000 on a Kaby Lake KVM guest with 8MB dentry hash and mitigations=off (performance is in the noise, under 1% difference, page tables are likely to be well cached for this workload). Similar numbers are seen on POWER9.

---
  include/asm-generic/4level-fixup.h |   1 +
  include/asm-generic/5level-fixup.h |   1 +
  include/linux/vmalloc.h            |   1 +
  mm/vmalloc.c                       | 132 +++++++++++++++++++++++------
  4 files changed, 107 insertions(+), 28 deletions(-)
include/asm-generic/4level-fixup.h +1 -0
index e3667c9a33a5..3cc65a4dd093 100644
@@ -20,6 +20,7 @@
  #define pud_none(pud)			0
  #define pud_bad(pud)			0
  #define pud_present(pud)		1
+#define pud_large(pud)			0
  #define pud_ERROR(pud)			do { } while (0)
  #define pud_clear(pud)			pgd_clear(pud)
  #define pud_val(pud)			pgd_val(pud)
include/asm-generic/5level-fixup.h +1 -0
index bb6cb347018c..c4377db09a4f 100644
@@ -22,6 +22,7 @@
  #define p4d_none(p4d)			0
  #define p4d_bad(p4d)			0
  #define p4d_present(p4d)		1
+#define p4d_large(p4d)			0
  #define p4d_ERROR(p4d)			do { } while (0)
  #define p4d_clear(p4d)			pgd_clear(p4d)
  #define p4d_val(p4d)			pgd_val(p4d)
include/linux/vmalloc.h +1 -0
index 812bea5866d6..4c92dc608928 100644
@@ -42,6 +42,7 @@ struct vm_struct {
  	unsigned long		size;
  	unsigned long		flags;
  	struct page		**pages;
+	unsigned int		page_shift;
  	unsigned int		nr_pages;
  	phys_addr_t		phys_addr;
  	const void		*caller;
mm/vmalloc.c +21 -0
index dd27cfb29b10..0cf8e861caeb 100644
@@ -36,6 +36,7 @@
  #include <linux/rbtree_augmented.h>
 
  #include <linux/uaccess.h>
+#include <asm/pgtable.h>
  #include <asm/tlbflush.h>
  #include <asm/shmparam.h>
 
@@ -440,6 +441,41 @@ static int vmap_pages_range(unsigned long start, unsigned long end,
  	return ret;
  }
 
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+				   pgprot_t prot, struct page **pages,
+				   unsigned int page_shift)
+{
+	unsigned long addr = start;
+	unsigned int i, nr = (end - start) >> (PAGE_SHIFT + page_shift);
+
+	for (i = 0; i < nr; i++) {
+		int err;
+
+		err = vmap_range_noflush(addr,
+					addr + (PAGE_SIZE << page_shift),
+					__pa(page_address(pages[i])), prot,
+					PAGE_SHIFT + page_shift);
+		if (err)
+			return err;
+
+		addr += PAGE_SIZE << page_shift;
+	}

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 473

Quoted 10 lines
mm/vmalloc.c +10 -0
+	flush_cache_vmap(start, end);
+
+	return nr;
+}
+#else
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+			   pgprot_t prot, struct page **pages,
+			   unsigned int page_shift)
+{
+	BUG_ON(page_shift != PAGE_SIZE);

Do we really need a BUG_ON() there ? What happens if this condition is true ?

Quoted 20 lines
mm/vmalloc.c +6 -1
+	return vmap_pages_range(start, end, prot, pages);
+}
+#endif
+
+
  int is_vmalloc_or_module_addr(const void *x)
  {
  	/*
@@ -462,7 +498,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
  {
  	unsigned long addr = (unsigned long) vmalloc_addr;
  	struct page *page = NULL;
-	pgd_t *pgd = pgd_offset_k(addr);
+	pgd_t *pgd;
  	p4d_t *p4d;
  	pud_t *pud;
  	pmd_t *pmd;
@@ -474,27 +510,38 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
  	 */
  	VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr));

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 520

Quoted 10 lines
mm/vmalloc.c +3 -1
 
+	pgd = pgd_offset_k(addr);
  	if (pgd_none(*pgd))
  		return NULL;
+
  	p4d = p4d_offset(pgd, addr);
  	if (p4d_none(*p4d))
  		return NULL;
-	pud = pud_offset(p4d, addr);
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP

Do we really need that ifdef ? Won't p4d_large() always return 0 when is not set ? Otherwise, could we use IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP) instead ?

Same several places below.

Quoted 73 lines
mm/vmalloc.c +27 -15
+	if (p4d_large(*p4d))
+		return p4d_page(*p4d) + ((addr & ~P4D_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(p4d_bad(*p4d)))
+		return NULL;
 
-	/*
-	 * Don't dereference bad PUD or PMD (below) entries. This will also
-	 * identify huge mappings, which we may encounter on architectures
-	 * that define CONFIG_HAVE_ARCH_HUGE_VMAP=y. Such regions will be
-	 * identified as vmalloc addresses by is_vmalloc_addr(), but are
-	 * not [unambiguously] associated with a struct page, so there is
-	 * no correct value to return for them.
-	 */
-	WARN_ON_ONCE(pud_bad(*pud));
-	if (pud_none(*pud) || pud_bad(*pud))
+	pud = pud_offset(p4d, addr);
+	if (pud_none(*pud))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pud_large(*pud))
+		return pud_page(*pud) + ((addr & ~PUD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pud_bad(*pud)))
  		return NULL;
+
  	pmd = pmd_offset(pud, addr);
-	WARN_ON_ONCE(pmd_bad(*pmd));
-	if (pmd_none(*pmd) || pmd_bad(*pmd))
+	if (pmd_none(*pmd))
+		return NULL;
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
+	if (pmd_large(*pmd))
+		return pmd_page(*pmd) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
+#endif
+	if (WARN_ON_ONCE(pmd_bad(*pmd)))
  		return NULL;
 
  	ptep = pte_offset_map(pmd, addr);
@@ -502,6 +549,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
  	if (pte_present(pte))
  		page = pte_page(pte);
  	pte_unmap(ptep);
+
  	return page;
  }
  EXPORT_SYMBOL(vmalloc_to_page);
@@ -2185,8 +2233,9 @@ static struct vm_struct *__get_vm_area_node(unsigned long size,
  		return NULL;
 
  	if (flags & VM_IOREMAP)
-		align = 1ul << clamp_t(int, get_count_order_long(size),
-				       PAGE_SHIFT, IOREMAP_MAX_ORDER);
+		align = max(align,
+				1ul << clamp_t(int, get_count_order_long(size),
+				       PAGE_SHIFT, IOREMAP_MAX_ORDER));
 
  	area = kzalloc_node(sizeof(*area), gfp_mask & GFP_RECLAIM_MASK, node);
  	if (unlikely(!area))
@@ -2398,7 +2447,7 @@ static void __vunmap(const void *addr, int deallocate_pages)
  			struct page *page = area->pages[i];
 
  			BUG_ON(!page);
-			__free_pages(page, 0);
+			__free_pages(page, area->page_shift);
  		}
 
  		kvfree(area->pages);
@@ -2541,14 +2590,17 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
  				 pgprot_t prot, int node)
  {
  	struct page **pages;
+	unsigned long addr = (unsigned long)area->addr;

mm/vmalloc.c static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, · line 2601

Quoted 10 lines
mm/vmalloc.c +4 -2
+	unsigned long size = get_vm_area_size(area);
+	unsigned int page_shift = area->page_shift;
+	unsigned int shift = page_shift + PAGE_SHIFT;
  	unsigned int nr_pages, array_size, i;
  	const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO;
  	const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN;
  	const gfp_t highmem_mask = (gfp_mask & (GFP_DMA | GFP_DMA32)) ?
-					0 :
-					__GFP_HIGHMEM;
+					0 : __GFP_HIGHMEM;

This patch is already quite big, shouldn't this kind of unrelated cleanups be in another patch ?

Quoted 7 lines
mm/vmalloc.c +1 -1
-	nr_pages = get_vm_area_size(area) >> PAGE_SHIFT;
+	nr_pages = size >> shift;
  	array_size = (nr_pages * sizeof(struct page *));
 
  	area->nr_pages = nr_pages;

mm/vmalloc.c static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, · line 2625

Quoted 9 lines
mm/vmalloc.c +2 -4
@@ -2569,10 +2621,8 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
  	for (i = 0; i < area->nr_pages; i++) {
  		struct page *page;
 
-		if (node == NUMA_NO_NODE)
-			page = alloc_page(alloc_mask|highmem_mask);
-		else
-			page = alloc_pages_node(node, alloc_mask|highmem_mask, 0);
+		page = alloc_pages_node(node,
+				alloc_mask|highmem_mask, page_shift);

This is also nice cleanup, but does it really belong to this patch ?

mm/vmalloc.c static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, · line 2639

Quoted 11 lines
mm/vmalloc.c +2 -1
  		if (unlikely(!page)) {
  			/* Successfully allocated i pages, free them in __vunmap() */
@@ -2584,8 +2634,9 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
  			cond_resched();
  	}
 
-	if (map_vm_area(area, prot, pages))
+	if (vmap_hpages_range(addr, addr + size, prot, pages, page_shift) < 0)
  		goto fail;
+

Cleanup ?

Quoted 54 lines
mm/vmalloc.c +22 -3
  	return area->addr;
 
  fail:
@@ -2619,22 +2670,39 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
  			pgprot_t prot, unsigned long vm_flags, int node,
  			const void *caller)
  {
-	struct vm_struct *area;
+	struct vm_struct *area = NULL;
  	void *addr;
  	unsigned long real_size = size;
+	unsigned long real_align = align;
+	unsigned int shift = PAGE_SHIFT;
 
  	size = PAGE_ALIGN(size);
  	if (!size || (size >> PAGE_SHIFT) > totalram_pages())
  		goto fail;
 
+	if (IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP)) {
+		unsigned long size_per_node;
+
+		size_per_node = size;
+		if (node == NUMA_NO_NODE)
+			size_per_node /= num_online_nodes();
+		if (size_per_node >= PMD_SIZE)
+			shift = PMD_SHIFT;
+	}
+again:
+	align = max(real_align, 1UL << shift);
+	size = ALIGN(real_size, align);
+
  	area = __get_vm_area_node(size, align, VM_ALLOC | VM_UNINITIALIZED |
  				vm_flags, start, end, node, gfp_mask, caller);
  	if (!area)
  		goto fail;
 
+	area->page_shift = shift - PAGE_SHIFT;
+
  	addr = __vmalloc_area_node(area, gfp_mask, prot, node);
  	if (!addr)
-		return NULL;
+		goto fail;
 
  	/*
  	 * In this function, newly allocated vm_struct has VM_UNINITIALIZED
@@ -2648,8 +2716,16 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align,
  	return addr;
 
  fail:
-	warn_alloc(gfp_mask, NULL,
+	if (shift == PMD_SHIFT) {
+		shift = PAGE_SHIFT;
+		goto again;

mm/vmalloc.c void *__vmalloc_node_range(unsigned long size, unsigned long align, · line 2731

Quoted 10 lines
mm/vmalloc.c +6 -0
+	}
+
+	if (!area) {
+		/* Warn for area allocation, page allocations already warn */
+		warn_alloc(gfp_mask, NULL,
  			  "vmalloc: allocation failure: %lu bytes", real_size);
+	}
  	return NULL;
  }
 

Christophe

author reviewer
#17 lore raw

Re: [PATCH 4/4] mm/vmalloc: Hugepage vmalloc mappings

Christophe Leroy's on June 11, 2019 3:39 pm:

Quoted 6 lines

Le 10/06/2019 à 06:38, Nicholas Piggin a écrit :

Quoted 2 lines

For platforms that define HAVE_ARCH_HUGE_VMAP, have vmap allow vmalloc to allocate huge pages and map them

Quoted 7 lines

Will this be compatible with Russell's series
https://patchwork.ozlabs.org/patch/1099857/ for the implementation of STRICT_MODULE_RWX ?
I see that apply_to_page_range() have things like BUG_ON(pud_huge(*pud));

Might also be an issue for arm64 as I think Russell's implementation comes from there.

Yeah you're right (and correct about arm64 problem). I'll fix that up.

Quoted 8 lines

mm/vmalloc.c static int vmap_pages_range(unsigned long start, unsigned long end, · line 473

Quoted 5 lines
mm/vmalloc.c +5 -0
+static int vmap_hpages_range(unsigned long start, unsigned long end,
+			   pgprot_t prot, struct page **pages,
+			   unsigned int page_shift)
+{
+	BUG_ON(page_shift != PAGE_SIZE);

Do we really need a BUG_ON() there ? What happens if this condition is true ?

If it's true then vmap_pages_range would die horribly reading off the end of the pages array thinking they are struct page pointers.

I guess it could return failure.

Quoted 35 lines
Quoted 20 lines
mm/vmalloc.c +6 -1
+	return vmap_pages_range(start, end, prot, pages);
+}
+#endif
+
+
  int is_vmalloc_or_module_addr(const void *x)
  {
  	/*
@@ -462,7 +498,7 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
  {
  	unsigned long addr = (unsigned long) vmalloc_addr;
  	struct page *page = NULL;
-	pgd_t *pgd = pgd_offset_k(addr);
+	pgd_t *pgd;
  	p4d_t *p4d;
  	pud_t *pud;
  	pmd_t *pmd;
@@ -474,27 +510,38 @@ struct page *vmalloc_to_page(const void *vmalloc_addr)
  	 */
  	VIRTUAL_BUG_ON(!is_vmalloc_or_module_addr(vmalloc_addr));

mm/vmalloc.c struct page *vmalloc_to_page(const void *vmalloc_addr) · line 520

Quoted 10 lines
mm/vmalloc.c +3 -1
 
+	pgd = pgd_offset_k(addr);
  	if (pgd_none(*pgd))
  		return NULL;
+
  	p4d = p4d_offset(pgd, addr);
  	if (p4d_none(*p4d))
  		return NULL;
-	pud = pud_offset(p4d, addr);
+#ifdef CONFIG_HAVE_ARCH_HUGE_VMAP

Do we really need that ifdef ? Won't p4d_large() always return 0 when is not set ?
Otherwise, could we use IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP) instead ?

Quoted 1 line

Same several places below.

Possibly some of them are not defined without HAVE_ARCH_HUGE_VMAP I think. I'll try to apply this pattern as much as possible.

Quoted 10 lines
Quoted 10 lines
mm/vmalloc.c +4 -0
@@ -2541,14 +2590,17 @@ static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask,
  				 pgprot_t prot, int node)
  {
  	struct page **pages;
+	unsigned long addr = (unsigned long)area->addr;
+	unsigned long size = get_vm_area_size(area);
+	unsigned int page_shift = area->page_shift;
+	unsigned int shift = page_shift + PAGE_SHIFT;
  	unsigned int nr_pages, array_size, i;
  	const gfp_t nested_gfp = (gfp_mask & GFP_RECLAIM_MASK) | __GFP_ZERO;
Quoted 8 lines

mm/vmalloc.c static void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, · line 2601

Quoted 5 lines
mm/vmalloc.c +1 -2
  	const gfp_t alloc_mask = gfp_mask | __GFP_NOWARN;
  	const gfp_t highmem_mask = (gfp_mask & (GFP_DMA | GFP_DMA32)) ?
-					0 :
-					__GFP_HIGHMEM;
+					0 : __GFP_HIGHMEM;

This patch is already quite big, shouldn't this kind of unrelated cleanups be in another patch ?

Okay, 2 against 1. I'll minimise changes like this.

Thanks,
Nick

#18 lore raw
In reply to Nicholas Piggin
Quoted 2 lines · Nicholas Piggin

ioremap_page_range is a generic function to create a kernel virtual mapping, move it to mm/vmalloc.c and rename it vmap_range.

Absolutely. It belongs in mm/vmalloc.c as its a kernel virtual range. But what is the rationale of changing the name to vmap_range ?

Quoted 3 lines

For clarity with this move, also:
- Rename vunmap_page_range (vmap_range's inverse) to vunmap_range.

Will be inverse for both vmap_range() and vmap_page[s]_range() ?

Quoted 1 line

- Rename vmap_page_range (which takes a page array) to vmap_pages.

s/vmap_pages/vmap_pages_range instead here ................^^^^^^

This deviates from the subject of this patch that it is related to ioremap only. I believe what this patch intends is to create

- vunmap_range() takes [VA range]

This will be the common kernel virtual range tear down function for ranges created either with vmap_range() or vmap_pages_range(). Is that correct ?

- vmap_range() takes [VA range, PA range, prot]
- vmap_pages_range() takes [VA range, struct pages, prot]

Can we re-order the arguments (pages <--> prot) for vmap_pages_range() just to make it sync with vmap_range() ?

static int vmap_pages_range(unsigned long start, unsigned long end,
pgprot_t prot, struct page **pages)

author reviewer
#19 lore raw

Anshuman Khandual's on June 10, 2019 3:42 pm:

Quoted 8 lines · Nicholas Piggin
Quoted 2 lines

ioremap_page_range is a generic function to create a kernel virtual mapping, move it to mm/vmalloc.c and rename it vmap_range.

Absolutely. It belongs in mm/vmalloc.c as its a kernel virtual range. But what is the rationale of changing the name to vmap_range ?

Well it doesn't just map IO. It's for arbitrary kernel virtual mapping (including ioremap). Last patch uses it to map regular cacheable memory.

Quoted 4 lines
Quoted 2 lines

For clarity with this move, also:
- Rename vunmap_page_range (vmap_range's inverse) to vunmap_range.

Will be inverse for both vmap_range() and vmap_page[s]_range() ?

Yes.

Quoted 4 lines
Quoted 2 lines

- Rename vmap_page_range (which takes a page array) to vmap_pages.

s/vmap_pages/vmap_pages_range instead here ................^^^^^^

Yes.

Quoted 10 lines

This deviates from the subject of this patch that it is related to ioremap only. I believe what this patch intends is to create

- vunmap_range() takes [VA range]

This will be the common kernel virtual range tear down function for ranges created either with vmap_range() or vmap_pages_range(). Is that correct ?
- vmap_range() takes [VA range, PA range, prot]
- vmap_pages_range() takes [VA range, struct pages, prot]

That's right although we already have all those functions, so I don't create anything, only move and re-name. I'm happy to change the subject if you have a preference.

Quoted 6 lines

Can we re-order the arguments (pages <--> prot) for vmap_pages_range() just to make it sync with vmap_range() ?

static int vmap_pages_range(unsigned long start, unsigned long end,
pgprot_t prot, struct page **pages)

Sure, makes sense.

Thanks,
Nick

#20 lore raw
In reply to Nicholas Piggin

Le 10/06/2019 à 06:38, Nicholas Piggin a écrit :

Quoted 6 lines

ioremap_page_range is a generic function to create a kernel virtual mapping, move it to mm/vmalloc.c and rename it vmap_range.

For clarity with this move, also:
- Rename vunmap_page_range (vmap_range's inverse) to vunmap_range.
- Rename vmap_page_range (which takes a page array) to vmap_pages.

Maybe it would be easier to follow the change if the name change was done in another patch than the move.

Quoted 20 lines

---

Fixed up the arm64 compile errors, fixed a few bugs, and tidied things up a bit more.

Have tested powerpc and x86 but not arm64, would appreciate a review and test of the arm64 patch if possible.

  include/linux/vmalloc.h |   3 +
  lib/ioremap.c           | 173 +++---------------------------
  mm/vmalloc.c            | 228 ++++++++++++++++++++++++++++++++++++----
  3 files changed, 229 insertions(+), 175 deletions(-)
include/linux/vmalloc.h +0 -0
index 51e131245379..812bea5866d6 100644

include/linux/vmalloc.h extern struct vm_struct *find_vm_area(const void *addr); · line 152

Quoted 6 lines
include/linux/vmalloc.h +3 -0
@@ -147,6 +147,9 @@ extern struct vm_struct *find_vm_area(const void *addr);
  extern int map_vm_area(struct vm_struct *area, pgprot_t prot,
  			struct page **pages);
  #ifdef CONFIG_MMU
+extern int vmap_range(unsigned long addr,
+		       unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+		       unsigned int max_page_shift);

Drop extern keyword here.

As checkpatch tells you, 'CHECK:AVOID_EXTERNS: extern prototypes should be avoided in .h files'

Christophe

Quoted 548 lines
include/linux/vmalloc.h +0 -0
  extern int map_kernel_range_noflush(unsigned long start, unsigned long size,
  				    pgprot_t prot, struct page **pages);
  extern void unmap_kernel_range_noflush(unsigned long addr, unsigned long size);
lib/ioremap.c +16 -157
index 063213685563..e13946da8ec3 100644
@@ -58,165 +58,24 @@ static inline int ioremap_pud_enabled(void) { return 0; }
  static inline int ioremap_pmd_enabled(void) { return 0; }
  #endif	/* CONFIG_HAVE_ARCH_HUGE_VMAP */
 
-static int ioremap_pte_range(pmd_t *pmd, unsigned long addr,
-		unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
-{
-	pte_t *pte;
-	u64 pfn;
-
-	pfn = phys_addr >> PAGE_SHIFT;
-	pte = pte_alloc_kernel(pmd, addr);
-	if (!pte)
-		return -ENOMEM;
-	do {
-		BUG_ON(!pte_none(*pte));
-		set_pte_at(&init_mm, addr, pte, pfn_pte(pfn, prot));
-		pfn++;
-	} while (pte++, addr += PAGE_SIZE, addr != end);
-	return 0;
-}
-
-static int ioremap_try_huge_pmd(pmd_t *pmd, unsigned long addr,
-				unsigned long end, phys_addr_t phys_addr,
-				pgprot_t prot)
-{
-	if (!ioremap_pmd_enabled())
-		return 0;
-
-	if ((end - addr) != PMD_SIZE)
-		return 0;
-
-	if (!IS_ALIGNED(phys_addr, PMD_SIZE))
-		return 0;
-
-	if (pmd_present(*pmd) && !pmd_free_pte_page(pmd, addr))
-		return 0;
-
-	return pmd_set_huge(pmd, phys_addr, prot);
-}
-
-static inline int ioremap_pmd_range(pud_t *pud, unsigned long addr,
-		unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
-{
-	pmd_t *pmd;
-	unsigned long next;
-
-	pmd = pmd_alloc(&init_mm, pud, addr);
-	if (!pmd)
-		return -ENOMEM;
-	do {
-		next = pmd_addr_end(addr, end);
-
-		if (ioremap_try_huge_pmd(pmd, addr, next, phys_addr, prot))
-			continue;
-
-		if (ioremap_pte_range(pmd, addr, next, phys_addr, prot))
-			return -ENOMEM;
-	} while (pmd++, phys_addr += (next - addr), addr = next, addr != end);
-	return 0;
-}
-
-static int ioremap_try_huge_pud(pud_t *pud, unsigned long addr,
-				unsigned long end, phys_addr_t phys_addr,
-				pgprot_t prot)
-{
-	if (!ioremap_pud_enabled())
-		return 0;
-
-	if ((end - addr) != PUD_SIZE)
-		return 0;
-
-	if (!IS_ALIGNED(phys_addr, PUD_SIZE))
-		return 0;
-
-	if (pud_present(*pud) && !pud_free_pmd_page(pud, addr))
-		return 0;
-
-	return pud_set_huge(pud, phys_addr, prot);
-}
-
-static inline int ioremap_pud_range(p4d_t *p4d, unsigned long addr,
-		unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
-{
-	pud_t *pud;
-	unsigned long next;
-
-	pud = pud_alloc(&init_mm, p4d, addr);
-	if (!pud)
-		return -ENOMEM;
-	do {
-		next = pud_addr_end(addr, end);
-
-		if (ioremap_try_huge_pud(pud, addr, next, phys_addr, prot))
-			continue;
-
-		if (ioremap_pmd_range(pud, addr, next, phys_addr, prot))
-			return -ENOMEM;
-	} while (pud++, phys_addr += (next - addr), addr = next, addr != end);
-	return 0;
-}
-
-static int ioremap_try_huge_p4d(p4d_t *p4d, unsigned long addr,
-				unsigned long end, phys_addr_t phys_addr,
-				pgprot_t prot)
-{
-	if (!ioremap_p4d_enabled())
-		return 0;
-
-	if ((end - addr) != P4D_SIZE)
-		return 0;
-
-	if (!IS_ALIGNED(phys_addr, P4D_SIZE))
-		return 0;
-
-	if (p4d_present(*p4d) && !p4d_free_pud_page(p4d, addr))
-		return 0;
-
-	return p4d_set_huge(p4d, phys_addr, prot);
-}
-
-static inline int ioremap_p4d_range(pgd_t *pgd, unsigned long addr,
-		unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
-{
-	p4d_t *p4d;
-	unsigned long next;
-
-	p4d = p4d_alloc(&init_mm, pgd, addr);
-	if (!p4d)
-		return -ENOMEM;
-	do {
-		next = p4d_addr_end(addr, end);
-
-		if (ioremap_try_huge_p4d(p4d, addr, next, phys_addr, prot))
-			continue;
-
-		if (ioremap_pud_range(p4d, addr, next, phys_addr, prot))
-			return -ENOMEM;
-	} while (p4d++, phys_addr += (next - addr), addr = next, addr != end);
-	return 0;
-}
-
  int ioremap_page_range(unsigned long addr,
  		       unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
  {
-	pgd_t *pgd;
-	unsigned long start;
-	unsigned long next;
-	int err;
-
-	might_sleep();
-	BUG_ON(addr >= end);
-
-	start = addr;
-	pgd = pgd_offset_k(addr);
-	do {
-		next = pgd_addr_end(addr, end);
-		err = ioremap_p4d_range(pgd, addr, next, phys_addr, prot);
-		if (err)
-			break;
-	} while (pgd++, phys_addr += (next - addr), addr = next, addr != end);
-
-	flush_cache_vmap(start, end);
+	unsigned int max_page_shift = PAGE_SHIFT;
+
+	/*
+	 * Due to the max_page_shift parameter to vmap_range, platforms must
+	 * enable all smaller sizes to take advantage of a given size,
+	 * otherwise fall back to small pages.
+	 */
+	if (ioremap_pmd_enabled()) {
+		max_page_shift = PMD_SHIFT;
+		if (ioremap_pud_enabled()) {
+			max_page_shift = PUD_SHIFT;
+			if (ioremap_p4d_enabled())
+				max_page_shift = P4D_SHIFT;
+		}
+	}
 
-	return err;
+	return vmap_range(addr, end, phys_addr, prot, max_page_shift);
  }
mm/vmalloc.c +210 -18
index 233af6936c93..dd27cfb29b10 100644
@@ -119,7 +119,7 @@ static void vunmap_p4d_range(pgd_t *pgd, unsigned long addr, unsigned long end)
  	} while (p4d++, addr = next, addr != end);
  }
 
-static void vunmap_page_range(unsigned long addr, unsigned long end)
+static void vunmap_range(unsigned long addr, unsigned long end)
  {
  	pgd_t *pgd;
  	unsigned long next;
@@ -135,6 +135,198 @@ static void vunmap_page_range(unsigned long addr, unsigned long end)
  }
 
  static int vmap_pte_range(pmd_t *pmd, unsigned long addr,
+			unsigned long end, phys_addr_t phys_addr, pgprot_t prot)
+{
+	pte_t *pte;
+	u64 pfn;
+
+	pfn = phys_addr >> PAGE_SHIFT;
+	pte = pte_alloc_kernel(pmd, addr);
+	if (!pte)
+		return -ENOMEM;
+	do {
+		BUG_ON(!pte_none(*pte));
+		set_pte_at(&init_mm, addr, pte, pfn_pte(pfn, prot));
+		pfn++;
+	} while (pte++, addr += PAGE_SIZE, addr != end);
+	return 0;
+}
+
+static int vmap_try_huge_pmd(pmd_t *pmd, unsigned long addr,
+			unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+			unsigned int max_page_shift)
+{
+	if (!IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP))
+		return 0;
+
+	if (max_page_shift < PMD_SHIFT)
+		return 0;
+
+	if ((end - addr) != PMD_SIZE)
+		return 0;
+
+	if (!IS_ALIGNED(phys_addr, PMD_SIZE))
+		return 0;
+
+	if (pmd_present(*pmd) && !pmd_free_pte_page(pmd, addr))
+		return 0;
+
+	return pmd_set_huge(pmd, phys_addr, prot);
+}
+
+static inline int vmap_pmd_range(pud_t *pud, unsigned long addr,
+			unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+			unsigned int max_page_shift)
+{
+	pmd_t *pmd;
+	unsigned long next;
+
+	pmd = pmd_alloc(&init_mm, pud, addr);
+	if (!pmd)
+		return -ENOMEM;
+	do {
+		next = pmd_addr_end(addr, end);
+
+		if (vmap_try_huge_pmd(pmd, addr, next, phys_addr, prot,
+					max_page_shift))
+			continue;
+
+		if (vmap_pte_range(pmd, addr, next, phys_addr, prot))
+			return -ENOMEM;
+	} while (pmd++, phys_addr += (next - addr), addr = next, addr != end);
+	return 0;
+}
+
+static int vmap_try_huge_pud(pud_t *pud, unsigned long addr,
+			unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+			unsigned int max_page_shift)
+{
+	if (!IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP))
+		return 0;
+
+	if (max_page_shift < PUD_SHIFT)
+		return 0;
+
+	if ((end - addr) != PUD_SIZE)
+		return 0;
+
+	if (!IS_ALIGNED(phys_addr, PUD_SIZE))
+		return 0;
+
+	if (pud_present(*pud) && !pud_free_pmd_page(pud, addr))
+		return 0;
+
+	return pud_set_huge(pud, phys_addr, prot);
+}
+
+static inline int vmap_pud_range(p4d_t *p4d, unsigned long addr,
+			unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+			unsigned int max_page_shift)
+{
+	pud_t *pud;
+	unsigned long next;
+
+	pud = pud_alloc(&init_mm, p4d, addr);
+	if (!pud)
+		return -ENOMEM;
+	do {
+		next = pud_addr_end(addr, end);
+
+		if (vmap_try_huge_pud(pud, addr, next, phys_addr, prot,
+					max_page_shift))
+			continue;
+
+		if (vmap_pmd_range(pud, addr, next, phys_addr, prot,
+					max_page_shift))
+			return -ENOMEM;
+	} while (pud++, phys_addr += (next - addr), addr = next, addr != end);
+	return 0;
+}
+
+static int vmap_try_huge_p4d(p4d_t *p4d, unsigned long addr,
+			unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+			unsigned int max_page_shift)
+{
+	if (!IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMAP))
+		return 0;
+
+	if (max_page_shift < P4D_SHIFT)
+		return 0;
+
+	if ((end - addr) != P4D_SIZE)
+		return 0;
+
+	if (!IS_ALIGNED(phys_addr, P4D_SIZE))
+		return 0;
+
+	if (p4d_present(*p4d) && !p4d_free_pud_page(p4d, addr))
+		return 0;
+
+	return p4d_set_huge(p4d, phys_addr, prot);
+}
+
+static inline int vmap_p4d_range(pgd_t *pgd, unsigned long addr,
+			unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+			unsigned int max_page_shift)
+{
+	p4d_t *p4d;
+	unsigned long next;
+
+	p4d = p4d_alloc(&init_mm, pgd, addr);
+	if (!p4d)
+		return -ENOMEM;
+	do {
+		next = p4d_addr_end(addr, end);
+
+		if (vmap_try_huge_p4d(p4d, addr, next, phys_addr, prot,
+					max_page_shift))
+			continue;
+
+		if (vmap_pud_range(p4d, addr, next, phys_addr, prot,
+					max_page_shift))
+			return -ENOMEM;
+	} while (p4d++, phys_addr += (next - addr), addr = next, addr != end);
+	return 0;
+}
+
+static int vmap_range_noflush(unsigned long addr,
+			unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+			unsigned int max_page_shift)
+{
+	pgd_t *pgd;
+	unsigned long start;
+	unsigned long next;
+	int err;
+
+	might_sleep();
+	BUG_ON(addr >= end);
+
+	start = addr;
+	pgd = pgd_offset_k(addr);
+	do {
+		next = pgd_addr_end(addr, end);
+		err = vmap_p4d_range(pgd, addr, next, phys_addr, prot,
+					max_page_shift);
+		if (err)
+			break;
+	} while (pgd++, phys_addr += (next - addr), addr = next, addr != end);
+
+	return err;
+}
+
+int vmap_range(unsigned long addr,
+		       unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+		       unsigned int max_page_shift)
+{
+	int ret;
+
+	ret = vmap_range_noflush(addr, end, phys_addr, prot, max_page_shift);
+	flush_cache_vmap(addr, end);
+
+	return ret;
+}
+
+static int vmap_pages_pte_range(pmd_t *pmd, unsigned long addr,
  		unsigned long end, pgprot_t prot, struct page **pages, int *nr)
  {
  	pte_t *pte;
@@ -160,7 +352,7 @@ static int vmap_pte_range(pmd_t *pmd, unsigned long addr,
  	return 0;
  }
 
-static int vmap_pmd_range(pud_t *pud, unsigned long addr,
+static int vmap_pages_pmd_range(pud_t *pud, unsigned long addr,
  		unsigned long end, pgprot_t prot, struct page **pages, int *nr)
  {
  	pmd_t *pmd;
@@ -171,13 +363,13 @@ static int vmap_pmd_range(pud_t *pud, unsigned long addr,
  		return -ENOMEM;
  	do {
  		next = pmd_addr_end(addr, end);
-		if (vmap_pte_range(pmd, addr, next, prot, pages, nr))
+		if (vmap_pages_pte_range(pmd, addr, next, prot, pages, nr))
  			return -ENOMEM;
  	} while (pmd++, addr = next, addr != end);
  	return 0;
  }
 
-static int vmap_pud_range(p4d_t *p4d, unsigned long addr,
+static int vmap_pages_pud_range(p4d_t *p4d, unsigned long addr,
  		unsigned long end, pgprot_t prot, struct page **pages, int *nr)
  {
  	pud_t *pud;
@@ -188,13 +380,13 @@ static int vmap_pud_range(p4d_t *p4d, unsigned long addr,
  		return -ENOMEM;
  	do {
  		next = pud_addr_end(addr, end);
-		if (vmap_pmd_range(pud, addr, next, prot, pages, nr))
+		if (vmap_pages_pmd_range(pud, addr, next, prot, pages, nr))
  			return -ENOMEM;
  	} while (pud++, addr = next, addr != end);
  	return 0;
  }
 
-static int vmap_p4d_range(pgd_t *pgd, unsigned long addr,
+static int vmap_pages_p4d_range(pgd_t *pgd, unsigned long addr,
  		unsigned long end, pgprot_t prot, struct page **pages, int *nr)
  {
  	p4d_t *p4d;
@@ -205,7 +397,7 @@ static int vmap_p4d_range(pgd_t *pgd, unsigned long addr,
  		return -ENOMEM;
  	do {
  		next = p4d_addr_end(addr, end);
-		if (vmap_pud_range(p4d, addr, next, prot, pages, nr))
+		if (vmap_pages_pud_range(p4d, addr, next, prot, pages, nr))
  			return -ENOMEM;
  	} while (p4d++, addr = next, addr != end);
  	return 0;
@@ -217,7 +409,7 @@ static int vmap_p4d_range(pgd_t *pgd, unsigned long addr,
   *
   * Ie. pte at addr+N*PAGE_SIZE shall point to pfn corresponding to pages[N]
   */
-static int vmap_page_range_noflush(unsigned long start, unsigned long end,
+static int vmap_pages_range_noflush(unsigned long start, unsigned long end,
  				   pgprot_t prot, struct page **pages)
  {
  	pgd_t *pgd;
@@ -230,7 +422,7 @@ static int vmap_page_range_noflush(unsigned long start, unsigned long end,
  	pgd = pgd_offset_k(addr);
  	do {
  		next = pgd_addr_end(addr, end);
-		err = vmap_p4d_range(pgd, addr, next, prot, pages, &nr);
+		err = vmap_pages_p4d_range(pgd, addr, next, prot, pages, &nr);
  		if (err)
  			return err;
  	} while (pgd++, addr = next, addr != end);
@@ -238,12 +430,12 @@ static int vmap_page_range_noflush(unsigned long start, unsigned long end,
  	return nr;
  }
 
-static int vmap_page_range(unsigned long start, unsigned long end,
+static int vmap_pages_range(unsigned long start, unsigned long end,
  			   pgprot_t prot, struct page **pages)
  {
  	int ret;
 
-	ret = vmap_page_range_noflush(start, end, prot, pages);
+	ret = vmap_pages_range_noflush(start, end, prot, pages);
  	flush_cache_vmap(start, end);
  	return ret;
  }
@@ -1148,7 +1340,7 @@ static void free_vmap_area(struct vmap_area *va)
   */
  static void unmap_vmap_area(struct vmap_area *va)
  {
-	vunmap_page_range(va->va_start, va->va_end);
+	vunmap_range(va->va_start, va->va_end);
  }
 
  /*
@@ -1586,7 +1778,7 @@ static void vb_free(const void *addr, unsigned long size)
  	rcu_read_unlock();
  	BUG_ON(!vb);
 
-	vunmap_page_range((unsigned long)addr, (unsigned long)addr + size);
+	vunmap_range((unsigned long)addr, (unsigned long)addr + size);
 
  	if (debug_pagealloc_enabled())
  		flush_tlb_kernel_range((unsigned long)addr,
@@ -1736,7 +1928,7 @@ void *vm_map_ram(struct page **pages, unsigned int count, int node, pgprot_t pro
  		addr = va->va_start;
  		mem = (void *)addr;
  	}
-	if (vmap_page_range(addr, addr + size, prot, pages) < 0) {
+	if (vmap_pages_range(addr, addr + size, prot, pages) < 0) {
  		vm_unmap_ram(mem, count);
  		return NULL;
  	}
@@ -1903,7 +2095,7 @@ void __init vmalloc_init(void)
  int map_kernel_range_noflush(unsigned long addr, unsigned long size,
  			     pgprot_t prot, struct page **pages)
  {
-	return vmap_page_range_noflush(addr, addr + size, prot, pages);
+	return vmap_pages_range_noflush(addr, addr + size, prot, pages);
  }
 
  /**
@@ -1922,7 +2114,7 @@ int map_kernel_range_noflush(unsigned long addr, unsigned long size,
   */
  void unmap_kernel_range_noflush(unsigned long addr, unsigned long size)
  {
-	vunmap_page_range(addr, addr + size);
+	vunmap_range(addr, addr + size);
  }
  EXPORT_SYMBOL_GPL(unmap_kernel_range_noflush);
 
@@ -1939,7 +2131,7 @@ void unmap_kernel_range(unsigned long addr, unsigned long size)
  	unsigned long end = addr + size;
 
  	flush_cache_vunmap(addr, end);
-	vunmap_page_range(addr, end);
+	vunmap_range(addr, end);
  	flush_tlb_kernel_range(addr, end);
  }
  EXPORT_SYMBOL_GPL(unmap_kernel_range);
@@ -1950,7 +2142,7 @@ int map_vm_area(struct vm_struct *area, pgprot_t prot, struct page **pages)
  	unsigned long end = addr + get_vm_area_size(area);
  	int err;
 
-	err = vmap_page_range(addr, end, prot, pages);
+	err = vmap_pages_range(addr, end, prot, pages);
 
  	return err > 0 ? 0 : err;
  }
author reviewer
#21 lore raw

Christophe Leroy's on June 11, 2019 3:24 pm:

Quoted 6 lines

Le 10/06/2019 à 06:38, Nicholas Piggin a écrit :

Quoted 3 lines

ioremap_page_range is a generic function to create a kernel virtual mapping, move it to mm/vmalloc.c and rename it vmap_range.

Quoted 6 lines
Quoted 3 lines

For clarity with this move, also:
- Rename vunmap_page_range (vmap_range's inverse) to vunmap_range.
- Rename vmap_page_range (which takes a page array) to vmap_pages.

Maybe it would be easier to follow the change if the name change was done in another patch than the move.

I could do that.

Quoted 14 lines
Quoted 14 lines

---

Fixed up the arm64 compile errors, fixed a few bugs, and tidied things up a bit more.

Have tested powerpc and x86 but not arm64, would appreciate a review and test of the arm64 patch if possible.

  include/linux/vmalloc.h |   3 +
  lib/ioremap.c           | 173 +++---------------------------
  mm/vmalloc.c            | 228 ++++++++++++++++++++++++++++++++++++----
  3 files changed, 229 insertions(+), 175 deletions(-)
Quoted 13 lines

include/linux/vmalloc.h extern struct vm_struct *find_vm_area(const void *addr); · line 152

Quoted 11 lines
include/linux/vmalloc.h +3 -0
index 51e131245379..812bea5866d6 100644
@@ -147,6 +147,9 @@ extern struct vm_struct *find_vm_area(const void *addr);
  extern int map_vm_area(struct vm_struct *area, pgprot_t prot,
  			struct page **pages);
  #ifdef CONFIG_MMU
+extern int vmap_range(unsigned long addr,
+		       unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+		       unsigned int max_page_shift);

Drop extern keyword here.

I don't know if I was going crazy but at one point I was getting duplicate symbol errors that were fixed by adding extern somewhere. Maybe sleep depravation. However...

Quoted 2 lines

As checkpatch tells you, 'CHECK:AVOID_EXTERNS: extern prototypes should be avoided in .h files'

I prefer to follow existing style in surrounding code at the expense of some checkpatch warnings. If somebody later wants to "fix" it that's fine.

Thanks,
Nick

#22 lore raw

Le 19/06/2019 à 05:43, Nicholas Piggin a écrit :

Quoted 4 lines

Christophe Leroy's on June 11, 2019 3:24 pm:

Quoted 3 lines

Le 10/06/2019 à 06:38, Nicholas Piggin a écrit :

[snip]

Quoted 14 lines
Quoted 5 lines
Quoted 5 lines
include/linux/vmalloc.h +0 -0
index 51e131245379..812bea5866d6 100644
@@ -147,6 +147,9 @@ extern struct vm_struct *find_vm_area(const void *addr);
Quoted 8 lines

include/linux/vmalloc.h extern struct vm_struct *find_vm_area(const void *addr); · line 152

Quoted 6 lines
include/linux/vmalloc.h +3 -0
   extern int map_vm_area(struct vm_struct *area, pgprot_t prot,
   			struct page **pages);
   #ifdef CONFIG_MMU
+extern int vmap_range(unsigned long addr,
+		       unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+		       unsigned int max_page_shift);

Drop extern keyword here.

Quoted 2 lines

I don't know if I was going crazy but at one point I was getting duplicate symbol errors that were fixed by adding extern somewhere.

probably not on a function name ...

Quoted 8 lines

Maybe sleep depravation. However...

Quoted 2 lines

As checkpatch tells you, 'CHECK:AVOID_EXTERNS: extern prototypes should be avoided in .h files'

I prefer to follow existing style in surrounding code at the expense of some checkpatch warnings. If somebody later wants to "fix" it that's fine.

I don't think that's fine to 'fix' later things that could be done right from the begining. 'Cosmetic only' fixes never happen because they are a nightmare for backports, and a shame for 'git blame'.

In some patches, you add cleanups to make the code look nicer, and here you have the opportunity to make the code nice from the begining and you prefer repeating the errors done in the past ? You're surprising me.

Christophe

Quoted 4 lines

Thanks,
Nick

author reviewer
#23 lore raw

Christophe Leroy's on June 19, 2019 11:18 pm:

Quoted 24 lines

Le 19/06/2019 à 05:43, Nicholas Piggin a écrit :

Quoted 4 lines

Christophe Leroy's on June 11, 2019 3:24 pm:

Quoted 3 lines

Le 10/06/2019 à 06:38, Nicholas Piggin a écrit :

[snip]

Quoted 14 lines
Quoted 13 lines

include/linux/vmalloc.h extern struct vm_struct *find_vm_area(const void *addr); · line 152

Quoted 11 lines
include/linux/vmalloc.h +3 -0
index 51e131245379..812bea5866d6 100644
@@ -147,6 +147,9 @@ extern struct vm_struct *find_vm_area(const void *addr);
   extern int map_vm_area(struct vm_struct *area, pgprot_t prot,
   			struct page **pages);
   #ifdef CONFIG_MMU
+extern int vmap_range(unsigned long addr,
+		       unsigned long end, phys_addr_t phys_addr, pgprot_t prot,
+		       unsigned int max_page_shift);

Drop extern keyword here.

Quoted 4 lines
Quoted 2 lines

I don't know if I was going crazy but at one point I was getting duplicate symbol errors that were fixed by adding extern somewhere.

probably not on a function name ...

I know it sounds crazy :P

Quoted 7 lines
Quoted 6 lines
Quoted 2 lines

As checkpatch tells you, 'CHECK:AVOID_EXTERNS: extern prototypes should be avoided in .h files'

I prefer to follow existing style in surrounding code at the expense of some checkpatch warnings. If somebody later wants to "fix" it that's fine.

Quoted 7 lines

I don't think that's fine to 'fix' later things that could be done right from the begining. 'Cosmetic only' fixes never happen because they are a nightmare for backports, and a shame for 'git blame'.

In some patches, you add cleanups to make the code look nicer, and here you have the opportunity to make the code nice from the begining and you prefer repeating the errors done in the past ? You're surprising me.

Well I never claimed to be consistent. I actually don't mind the extern keyword so it's probably just my personal preference that makes me notice something nearby. I have dropped those "cleanup" changes though, so there.

Thanks,
Nick