linux/arch/powerpc/mm/hugetlbpage.c

919 lines
22 KiB
C
Raw Normal View History

/*
* PPC Huge TLB Page Support for Kernel.
*
* Copyright (C) 2003 David Gibson, IBM Corporation.
* Copyright (C) 2011 Becky Bruce, Freescale Semiconductor
*
* Based on the IA-32 version:
* Copyright (C) 2002, Rohit Seth <rohit.seth@intel.com>
*/
#include <linux/mm.h>
#include <linux/io.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
#include <linux/hugetlb.h>
#include <linux/of_fdt.h>
#include <linux/memblock.h>
#include <linux/bootmem.h>
#include <linux/moduleparam.h>
#include <asm/pgtable.h>
#include <asm/pgalloc.h>
#include <asm/tlb.h>
#include <asm/setup.h>
#define PAGE_SHIFT_64K 16
#define PAGE_SHIFT_16M 24
#define PAGE_SHIFT_16G 34
unsigned int HPAGE_SHIFT;
/*
* Tracks gpages after the device tree is scanned and before the
* huge_boot_pages list is ready. On non-Freescale implementations, this is
* just used to track 16G pages and so is a single array. FSL-based
* implementations may have more than one gpage size, so we need multiple
* arrays
*/
#ifdef CONFIG_PPC_FSL_BOOK3E
#define MAX_NUMBER_GPAGES 128
struct psize_gpages {
u64 gpage_list[MAX_NUMBER_GPAGES];
unsigned int nr_gpages;
};
static struct psize_gpages gpage_freearray[MMU_PAGE_COUNT];
#else
#define MAX_NUMBER_GPAGES 1024
static u64 gpage_freearray[MAX_NUMBER_GPAGES];
static unsigned nr_gpages;
#endif
static inline int shift_to_mmu_psize(unsigned int shift)
{
int psize;
for (psize = 0; psize < MMU_PAGE_COUNT; ++psize)
if (mmu_psize_defs[psize].shift == shift)
return psize;
return -1;
}
static inline unsigned int mmu_psize_to_shift(unsigned int mmu_psize)
{
if (mmu_psize_defs[mmu_psize].shift)
return mmu_psize_defs[mmu_psize].shift;
BUG();
}
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
#define hugepd_none(hpd) ((hpd).pd == 0)
pte_t *find_linux_pte_or_hugepte(pgd_t *pgdir, unsigned long ea, unsigned *shift)
{
pgd_t *pg;
pud_t *pu;
pmd_t *pm;
hugepd_t *hpdp = NULL;
unsigned pdshift = PGDIR_SHIFT;
if (shift)
*shift = 0;
pg = pgdir + pgd_index(ea);
if (is_hugepd(pg)) {
hpdp = (hugepd_t *)pg;
} else if (!pgd_none(*pg)) {
pdshift = PUD_SHIFT;
pu = pud_offset(pg, ea);
if (is_hugepd(pu))
hpdp = (hugepd_t *)pu;
else if (!pud_none(*pu)) {
pdshift = PMD_SHIFT;
pm = pmd_offset(pu, ea);
if (is_hugepd(pm))
hpdp = (hugepd_t *)pm;
else if (!pmd_none(*pm)) {
return pte_offset_kernel(pm, ea);
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
}
}
}
if (!hpdp)
return NULL;
if (shift)
*shift = hugepd_shift(*hpdp);
return hugepte_offset(hpdp, ea, pdshift);
}
pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr)
{
return find_linux_pte_or_hugepte(mm->pgd, addr, NULL);
}
static int __hugepte_alloc(struct mm_struct *mm, hugepd_t *hpdp,
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
unsigned long address, unsigned pdshift, unsigned pshift)
{
struct kmem_cache *cachep;
pte_t *new;
#ifdef CONFIG_PPC_FSL_BOOK3E
int i;
int num_hugepd = 1 << (pshift - pdshift);
cachep = hugepte_cache;
#else
cachep = PGT_CACHE(pdshift - pshift);
#endif
new = kmem_cache_zalloc(cachep, GFP_KERNEL|__GFP_REPEAT);
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
BUG_ON(pshift > HUGEPD_SHIFT_MASK);
BUG_ON((unsigned long)new & HUGEPD_SHIFT_MASK);
if (! new)
return -ENOMEM;
spin_lock(&mm->page_table_lock);
#ifdef CONFIG_PPC_FSL_BOOK3E
/*
* We have multiple higher-level entries that point to the same
* actual pte location. Fill in each as we go and backtrack on error.
* We need all of these so the DTLB pgtable walk code can find the
* right higher-level entry without knowing if it's a hugepage or not.
*/
for (i = 0; i < num_hugepd; i++, hpdp++) {
if (unlikely(!hugepd_none(*hpdp)))
break;
else
hpdp->pd = ((unsigned long)new & ~PD_HUGE) | pshift;
}
/* If we bailed from the for loop early, an error occurred, clean up */
if (i < num_hugepd) {
for (i = i - 1 ; i >= 0; i--, hpdp--)
hpdp->pd = 0;
kmem_cache_free(cachep, new);
}
#else
if (!hugepd_none(*hpdp))
kmem_cache_free(cachep, new);
else
hpdp->pd = ((unsigned long)new & ~PD_HUGE) | pshift;
#endif
spin_unlock(&mm->page_table_lock);
return 0;
}
/*
* These macros define how to determine which level of the page table holds
* the hpdp.
*/
#ifdef CONFIG_PPC_FSL_BOOK3E
#define HUGEPD_PGD_SHIFT PGDIR_SHIFT
#define HUGEPD_PUD_SHIFT PUD_SHIFT
#else
#define HUGEPD_PGD_SHIFT PUD_SHIFT
#define HUGEPD_PUD_SHIFT PMD_SHIFT
#endif
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
pte_t *huge_pte_alloc(struct mm_struct *mm, unsigned long addr, unsigned long sz)
{
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
pgd_t *pg;
pud_t *pu;
pmd_t *pm;
hugepd_t *hpdp = NULL;
unsigned pshift = __ffs(sz);
unsigned pdshift = PGDIR_SHIFT;
addr &= ~(sz-1);
pg = pgd_offset(mm, addr);
if (pshift >= HUGEPD_PGD_SHIFT) {
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
hpdp = (hugepd_t *)pg;
} else {
pdshift = PUD_SHIFT;
pu = pud_alloc(mm, pg, addr);
if (pshift >= HUGEPD_PUD_SHIFT) {
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
hpdp = (hugepd_t *)pu;
} else {
pdshift = PMD_SHIFT;
pm = pmd_alloc(mm, pu, addr);
hpdp = (hugepd_t *)pm;
}
}
if (!hpdp)
return NULL;
BUG_ON(!hugepd_none(*hpdp) && !hugepd_ok(*hpdp));
if (hugepd_none(*hpdp) && __hugepte_alloc(mm, hpdp, addr, pdshift, pshift))
return NULL;
return hugepte_offset(hpdp, addr, pdshift);
}
#ifdef CONFIG_PPC_FSL_BOOK3E
/* Build list of addresses of gigantic pages. This function is used in early
* boot before the buddy or bootmem allocator is setup.
*/
void add_gpage(u64 addr, u64 page_size, unsigned long number_of_pages)
{
unsigned int idx = shift_to_mmu_psize(__ffs(page_size));
int i;
if (addr == 0)
return;
gpage_freearray[idx].nr_gpages = number_of_pages;
for (i = 0; i < number_of_pages; i++) {
gpage_freearray[idx].gpage_list[i] = addr;
addr += page_size;
}
}
/*
* Moves the gigantic page addresses from the temporary list to the
* huge_boot_pages list.
*/
int alloc_bootmem_huge_page(struct hstate *hstate)
{
struct huge_bootmem_page *m;
int idx = shift_to_mmu_psize(hstate->order + PAGE_SHIFT);
int nr_gpages = gpage_freearray[idx].nr_gpages;
if (nr_gpages == 0)
return 0;
#ifdef CONFIG_HIGHMEM
/*
* If gpages can be in highmem we can't use the trick of storing the
* data structure in the page; allocate space for this
*/
m = alloc_bootmem(sizeof(struct huge_bootmem_page));
m->phys = gpage_freearray[idx].gpage_list[--nr_gpages];
#else
m = phys_to_virt(gpage_freearray[idx].gpage_list[--nr_gpages]);
#endif
list_add(&m->list, &huge_boot_pages);
gpage_freearray[idx].nr_gpages = nr_gpages;
gpage_freearray[idx].gpage_list[nr_gpages] = 0;
m->hstate = hstate;
return 1;
}
/*
* Scan the command line hugepagesz= options for gigantic pages; store those in
* a list that we use to allocate the memory once all options are parsed.
*/
unsigned long gpage_npages[MMU_PAGE_COUNT];
static int __init do_gpage_early_setup(char *param, char *val)
{
static phys_addr_t size;
unsigned long npages;
/*
* The hugepagesz and hugepages cmdline options are interleaved. We
* use the size variable to keep track of whether or not this was done
* properly and skip over instances where it is incorrect. Other
* command-line parsing code will issue warnings, so we don't need to.
*
*/
if ((strcmp(param, "default_hugepagesz") == 0) ||
(strcmp(param, "hugepagesz") == 0)) {
size = memparse(val, NULL);
} else if (strcmp(param, "hugepages") == 0) {
if (size != 0) {
if (sscanf(val, "%lu", &npages) <= 0)
npages = 0;
gpage_npages[shift_to_mmu_psize(__ffs(size))] = npages;
size = 0;
}
}
return 0;
}
/*
* This function allocates physical space for pages that are larger than the
* buddy allocator can handle. We want to allocate these in highmem because
* the amount of lowmem is limited. This means that this function MUST be
* called before lowmem_end_addr is set up in MMU_init() in order for the lmb
* allocate to grab highmem.
*/
void __init reserve_hugetlb_gpages(void)
{
static __initdata char cmdline[COMMAND_LINE_SIZE];
phys_addr_t size, base;
int i;
strlcpy(cmdline, boot_command_line, COMMAND_LINE_SIZE);
parse_args("hugetlb gpages", cmdline, NULL, 0, &do_gpage_early_setup);
/*
* Walk gpage list in reverse, allocating larger page sizes first.
* Skip over unsupported sizes, or sizes that have 0 gpages allocated.
* When we reach the point in the list where pages are no longer
* considered gpages, we're done.
*/
for (i = MMU_PAGE_COUNT-1; i >= 0; i--) {
if (mmu_psize_defs[i].shift == 0 || gpage_npages[i] == 0)
continue;
else if (mmu_psize_to_shift(i) < (MAX_ORDER + PAGE_SHIFT))
break;
size = (phys_addr_t)(1ULL << mmu_psize_to_shift(i));
base = memblock_alloc_base(size * gpage_npages[i], size,
MEMBLOCK_ALLOC_ANYWHERE);
add_gpage(base, size, gpage_npages[i]);
}
}
#else /* !PPC_FSL_BOOK3E */
/* Build list of addresses of gigantic pages. This function is used in early
* boot before the buddy or bootmem allocator is setup.
*/
void add_gpage(u64 addr, u64 page_size, unsigned long number_of_pages)
{
if (!addr)
return;
while (number_of_pages > 0) {
gpage_freearray[nr_gpages] = addr;
nr_gpages++;
number_of_pages--;
addr += page_size;
}
}
/* Moves the gigantic page addresses from the temporary list to the
* huge_boot_pages list.
*/
int alloc_bootmem_huge_page(struct hstate *hstate)
{
struct huge_bootmem_page *m;
if (nr_gpages == 0)
return 0;
m = phys_to_virt(gpage_freearray[--nr_gpages]);
gpage_freearray[nr_gpages] = 0;
list_add(&m->list, &huge_boot_pages);
m->hstate = hstate;
return 1;
}
#endif
[PATCH] shared page table for hugetlb page Following up with the work on shared page table done by Dave McCracken. This set of patch target shared page table for hugetlb memory only. The shared page table is particular useful in the situation of large number of independent processes sharing large shared memory segments. In the normal page case, the amount of memory saved from process' page table is quite significant. For hugetlb, the saving on page table memory is not the primary objective (as hugetlb itself already cuts down page table overhead significantly), instead, the purpose of using shared page table on hugetlb is to allow faster TLB refill and smaller cache pollution upon TLB miss. With PT sharing, pte entries are shared among hundreds of processes, the cache consumption used by all the page table is smaller and in return, application gets much higher cache hit ratio. One other effect is that cache hit ratio with hardware page walker hitting on pte in cache will be higher and this helps to reduce tlb miss latency. These two effects contribute to higher application performance. Signed-off-by: Ken Chen <kenneth.w.chen@intel.com> Acked-by: Hugh Dickins <hugh@veritas.com> Cc: Dave McCracken <dmccr@us.ibm.com> Cc: William Lee Irwin III <wli@holomorphy.com> Cc: "Luck, Tony" <tony.luck@intel.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Gibson <david@gibson.dropbear.id.au> Cc: Adam Litke <agl@us.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: "David S. Miller" <davem@davemloft.net> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-07 04:32:03 +00:00
int huge_pmd_unshare(struct mm_struct *mm, unsigned long *addr, pte_t *ptep)
{
return 0;
}
#ifdef CONFIG_PPC_FSL_BOOK3E
#define HUGEPD_FREELIST_SIZE \
((PAGE_SIZE - sizeof(struct hugepd_freelist)) / sizeof(pte_t))
struct hugepd_freelist {
struct rcu_head rcu;
unsigned int index;
void *ptes[0];
};
static DEFINE_PER_CPU(struct hugepd_freelist *, hugepd_freelist_cur);
static void hugepd_free_rcu_callback(struct rcu_head *head)
{
struct hugepd_freelist *batch =
container_of(head, struct hugepd_freelist, rcu);
unsigned int i;
for (i = 0; i < batch->index; i++)
kmem_cache_free(hugepte_cache, batch->ptes[i]);
free_page((unsigned long)batch);
}
static void hugepd_free(struct mmu_gather *tlb, void *hugepte)
{
struct hugepd_freelist **batchp;
batchp = &__get_cpu_var(hugepd_freelist_cur);
if (atomic_read(&tlb->mm->mm_users) < 2 ||
cpumask_equal(mm_cpumask(tlb->mm),
cpumask_of(smp_processor_id()))) {
kmem_cache_free(hugepte_cache, hugepte);
return;
}
if (*batchp == NULL) {
*batchp = (struct hugepd_freelist *)__get_free_page(GFP_ATOMIC);
(*batchp)->index = 0;
}
(*batchp)->ptes[(*batchp)->index++] = hugepte;
if ((*batchp)->index == HUGEPD_FREELIST_SIZE) {
call_rcu_sched(&(*batchp)->rcu, hugepd_free_rcu_callback);
*batchp = NULL;
}
}
#endif
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
static void free_hugepd_range(struct mmu_gather *tlb, hugepd_t *hpdp, int pdshift,
unsigned long start, unsigned long end,
unsigned long floor, unsigned long ceiling)
{
pte_t *hugepte = hugepd_page(*hpdp);
int i;
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
unsigned long pdmask = ~((1UL << pdshift) - 1);
unsigned int num_hugepd = 1;
#ifdef CONFIG_PPC_FSL_BOOK3E
/* Note: On fsl the hpdp may be the first of several */
num_hugepd = (1 << (hugepd_shift(*hpdp) - pdshift));
#else
unsigned int shift = hugepd_shift(*hpdp);
#endif
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
start &= pdmask;
if (start < floor)
return;
if (ceiling) {
ceiling &= pdmask;
if (! ceiling)
return;
}
if (end - 1 > ceiling - 1)
return;
for (i = 0; i < num_hugepd; i++, hpdp++)
hpdp->pd = 0;
tlb->need_flush = 1;
#ifdef CONFIG_PPC_FSL_BOOK3E
hugepd_free(tlb, hugepte);
#else
pgtable_free_tlb(tlb, hugepte, pdshift - shift);
#endif
}
static void hugetlb_free_pmd_range(struct mmu_gather *tlb, pud_t *pud,
unsigned long addr, unsigned long end,
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
unsigned long floor, unsigned long ceiling)
{
pmd_t *pmd;
unsigned long next;
unsigned long start;
start = addr;
do {
pmd = pmd_offset(pud, addr);
next = pmd_addr_end(addr, end);
if (pmd_none(*pmd))
continue;
#ifdef CONFIG_PPC_FSL_BOOK3E
/*
* Increment next by the size of the huge mapping since
* there may be more than one entry at this level for a
* single hugepage, but all of them point to
* the same kmem cache that holds the hugepte.
*/
next = addr + (1 << hugepd_shift(*(hugepd_t *)pmd));
#endif
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
free_hugepd_range(tlb, (hugepd_t *)pmd, PMD_SHIFT,
addr, next, floor, ceiling);
} while (addr = next, addr != end);
start &= PUD_MASK;
if (start < floor)
return;
if (ceiling) {
ceiling &= PUD_MASK;
if (!ceiling)
return;
}
if (end - 1 > ceiling - 1)
return;
pmd = pmd_offset(pud, start);
pud_clear(pud);
pmd_free_tlb(tlb, pmd, start);
}
static void hugetlb_free_pud_range(struct mmu_gather *tlb, pgd_t *pgd,
unsigned long addr, unsigned long end,
unsigned long floor, unsigned long ceiling)
{
pud_t *pud;
unsigned long next;
unsigned long start;
start = addr;
do {
pud = pud_offset(pgd, addr);
next = pud_addr_end(addr, end);
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
if (!is_hugepd(pud)) {
if (pud_none_or_clear_bad(pud))
continue;
hugetlb_free_pmd_range(tlb, pud, addr, next, floor,
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
ceiling);
} else {
#ifdef CONFIG_PPC_FSL_BOOK3E
/*
* Increment next by the size of the huge mapping since
* there may be more than one entry at this level for a
* single hugepage, but all of them point to
* the same kmem cache that holds the hugepte.
*/
next = addr + (1 << hugepd_shift(*(hugepd_t *)pud));
#endif
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
free_hugepd_range(tlb, (hugepd_t *)pud, PUD_SHIFT,
addr, next, floor, ceiling);
}
} while (addr = next, addr != end);
start &= PGDIR_MASK;
if (start < floor)
return;
if (ceiling) {
ceiling &= PGDIR_MASK;
if (!ceiling)
return;
}
if (end - 1 > ceiling - 1)
return;
pud = pud_offset(pgd, start);
pgd_clear(pgd);
pud_free_tlb(tlb, pud, start);
}
/*
* This function frees user-level page tables of a process.
*
* Must be called with pagetable lock held.
*/
void hugetlb_free_pgd_range(struct mmu_gather *tlb,
unsigned long addr, unsigned long end,
unsigned long floor, unsigned long ceiling)
{
pgd_t *pgd;
unsigned long next;
/*
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
* Because there are a number of different possible pagetable
* layouts for hugepage ranges, we limit knowledge of how
* things should be laid out to the allocation path
* (huge_pte_alloc(), above). Everything else works out the
* structure as it goes from information in the hugepd
* pointers. That means that we can't here use the
* optimization used in the normal page free_pgd_range(), of
* checking whether we're actually covering a large enough
* range to have to do anything at the top level of the walk
* instead of at the bottom.
*
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
* To make sense of this, you should probably go read the big
* block comment at the top of the normal free_pgd_range(),
* too.
*/
do {
next = pgd_addr_end(addr, end);
pgd = pgd_offset(tlb->mm, addr);
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
if (!is_hugepd(pgd)) {
if (pgd_none_or_clear_bad(pgd))
continue;
hugetlb_free_pud_range(tlb, pgd, addr, next, floor, ceiling);
} else {
#ifdef CONFIG_PPC_FSL_BOOK3E
/*
* Increment next by the size of the huge mapping since
* there may be more than one entry at the pgd level
* for a single hugepage, but all of them point to the
* same kmem cache that holds the hugepte.
*/
next = addr + (1 << hugepd_shift(*(hugepd_t *)pgd));
#endif
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
free_hugepd_range(tlb, (hugepd_t *)pgd, PGDIR_SHIFT,
addr, next, floor, ceiling);
}
} while (addr = next, addr != end);
}
struct page *
follow_huge_addr(struct mm_struct *mm, unsigned long address, int write)
{
pte_t *ptep;
struct page *page;
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
unsigned shift;
unsigned long mask;
ptep = find_linux_pte_or_hugepte(mm->pgd, address, &shift);
/* Verify it is a huge page else bail. */
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
if (!ptep || !shift)
return ERR_PTR(-EINVAL);
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
mask = (1UL << shift) - 1;
page = pte_page(*ptep);
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
if (page)
page += (address & mask) / PAGE_SIZE;
return page;
}
int pmd_huge(pmd_t pmd)
{
return 0;
}
int pud_huge(pud_t pud)
{
return 0;
}
struct page *
follow_huge_pmd(struct mm_struct *mm, unsigned long address,
pmd_t *pmd, int write)
{
BUG();
return NULL;
}
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
static noinline int gup_hugepte(pte_t *ptep, unsigned long sz, unsigned long addr,
unsigned long end, int write, struct page **pages, int *nr)
{
unsigned long mask;
unsigned long pte_end;
struct page *head, *page, *tail;
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
pte_t pte;
int refs;
pte_end = (addr + sz) & ~(sz-1);
if (pte_end < end)
end = pte_end;
pte = *ptep;
mask = _PAGE_PRESENT | _PAGE_USER;
if (write)
mask |= _PAGE_RW;
if ((pte_val(pte) & mask) != mask)
return 0;
/* hugepages are never "special" */
VM_BUG_ON(!pfn_valid(pte_pfn(pte)));
refs = 0;
head = pte_page(pte);
page = head + ((addr & (sz-1)) >> PAGE_SHIFT);
tail = page;
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
do {
VM_BUG_ON(compound_head(page) != head);
pages[*nr] = page;
(*nr)++;
page++;
refs++;
} while (addr += PAGE_SIZE, addr != end);
if (!page_cache_add_speculative(head, refs)) {
*nr -= refs;
return 0;
}
if (unlikely(pte_val(pte) != pte_val(*ptep))) {
/* Could be optimized better */
*nr -= refs;
while (refs--)
powerpc: get_hugepte() don't put_page() the wrong page "page" may have changed to point to the next hugepage after the loop completed, The references have been taken on the head page, so the put_page must happen there too. This is a longstanding issue pre-thp inclusion. It's totally unclear how these page_cache_add_speculative and pte_val(pte) != pte_val(*ptep) checks are necessary across all the powerpc gup_fast code, when x86 doesn't need any of that: there's no way the page can be freed with irq disabled so we're guaranteed the atomic_inc will happen on a page with page_count > 0 (so not needing the speculative check). The pte check is also meaningless on x86: no need to rollback on x86 if the pte changed, because the pte can still change a CPU tick after the check succeeded and it won't be rolled back in that case. The important thing is we got a reference on a valid page that was mapped there a CPU tick ago. So not knowing the soft tlb refill code of ppc64 in great detail I'm not removing the "speculative" page_count increase and the pte checks across all the code, but unless there's a strong reason for it they should be later cleaned up too. If a pte can change from huge to non-huge (like it could happen with THP) passing a pte_t *ptep to gup_hugepte() would also require to repeat the is_hugepd in gup_hugepte(), but that shouldn't happen with hugetlbfs only so I'm not altering that. Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Hugh Dickins <hughd@google.com> Cc: Johannes Weiner <jweiner@redhat.com> Cc: Rik van Riel <riel@redhat.com> Cc: Mel Gorman <mgorman@suse.de> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Acked-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-11-02 20:37:08 +00:00
put_page(head);
return 0;
}
/*
* Any tail page need their mapcount reference taken before we
* return.
*/
while (refs--) {
if (PageTail(tail))
get_huge_page_tail(tail);
tail++;
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
}
return 1;
}
static unsigned long hugepte_addr_end(unsigned long addr, unsigned long end,
unsigned long sz)
{
unsigned long __boundary = (addr + sz) & ~(sz-1);
return (__boundary - 1 < end - 1) ? __boundary : end;
}
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
int gup_hugepd(hugepd_t *hugepd, unsigned pdshift,
unsigned long addr, unsigned long end,
int write, struct page **pages, int *nr)
{
pte_t *ptep;
unsigned long sz = 1UL << hugepd_shift(*hugepd);
unsigned long next;
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
ptep = hugepte_offset(hugepd, addr, pdshift);
do {
next = hugepte_addr_end(addr, end, sz);
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
if (!gup_hugepte(ptep, sz, addr, end, write, pages, nr))
return 0;
} while (ptep++, addr = next, addr != end);
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
return 1;
}
#ifdef CONFIG_PPC_MM_SLICES
unsigned long hugetlb_get_unmapped_area(struct file *file, unsigned long addr,
unsigned long len, unsigned long pgoff,
unsigned long flags)
{
struct hstate *hstate = hstate_file(file);
int mmu_psize = shift_to_mmu_psize(huge_page_shift(hstate));
powerpc: Check for valid hugepage size in hugetlb_get_unmapped_area It looks like most of the hugetlb code is doing the correct thing if hugepages are not supported, but the mmap code is not. If we get into the mmap code when hugepages are not supported, such as in an LPAR which is running Active Memory Sharing, we can oops the kernel. This fixes the oops being seen in this path. oops: Kernel access of bad area, sig: 11 [#1] SMP NR_CPUS=1024 NUMA pSeries Modules linked in: nfs(N) lockd(N) nfs_acl(N) sunrpc(N) ipv6(N) fuse(N) loop(N) dm_mod(N) sg(N) ibmveth(N) sd_mod(N) crc_t10dif(N) ibmvscsic(N) scsi_transport_srp(N) scsi_tgt(N) scsi_mod(N) Supported: No NIP: c000000000038d60 LR: c00000000003945c CTR: c0000000000393f0 REGS: c000000077e7b830 TRAP: 0300 Tainted: G (2.6.27.5-bz50170-2-ppc64) MSR: 8000000000009032 <EE,ME,IR,DR> CR: 44000448 XER: 20000001 DAR: c000002000af90a8, DSISR: 0000000040000000 TASK = c00000007c1b8600[4019] 'hugemmap01' THREAD: c000000077e78000 CPU: 6 GPR00: 0000001fffffffe0 c000000077e7bab0 c0000000009a4e78 0000000000000000 GPR04: 0000000000010000 0000000000000001 00000000ffffffff 0000000000000001 GPR08: 0000000000000000 c000000000af90c8 0000000000000001 0000000000000000 GPR12: 000000000000003f c000000000a73880 0000000000000000 0000000000000000 GPR16: 0000000000000000 0000000000000000 0000000000000000 0000000000010000 GPR20: 0000000000000000 0000000000000003 0000000000010000 0000000000000001 GPR24: 0000000000000003 0000000000000000 0000000000000001 ffffffffffffffb5 GPR28: c000000077ca2e80 0000000000000000 c00000000092af78 0000000000010000 NIP [c000000000038d60] .slice_get_unmapped_area+0x6c/0x4e0 LR [c00000000003945c] .hugetlb_get_unmapped_area+0x6c/0x80 Call Trace: [c000000077e7bbc0] [c00000000003945c] .hugetlb_get_unmapped_area+0x6c/0x80 [c000000077e7bc30] [c000000000107e30] .get_unmapped_area+0x64/0xd8 [c000000077e7bcb0] [c00000000010b140] .do_mmap_pgoff+0x140/0x420 [c000000077e7bd80] [c00000000000bf5c] .sys_mmap+0xc4/0x140 [c000000077e7be30] [c0000000000086b4] syscall_exit+0x0/0x40 Instruction dump: fac1ffb0 fae1ffb8 fb01ffc0 fb21ffc8 fb41ffd0 fb61ffd8 fb81ffe0 fbc1fff0 fbe1fff8 f821fef1 f8c10158 f8e10160 <7d49002e> f9010168 e92d01b0 eb4902b0 Signed-off-by: Brian King <brking@linux.vnet.ibm.com> Signed-off-by: Paul Mackerras <paulus@samba.org>
2008-12-04 04:07:54 +00:00
return slice_get_unmapped_area(addr, len, flags, mmu_psize, 1, 0);
}
#endif
unsigned long vma_mmu_pagesize(struct vm_area_struct *vma)
{
#ifdef CONFIG_PPC_MM_SLICES
unsigned int psize = get_slice_psize(vma->vm_mm, vma->vm_start);
return 1UL << mmu_psize_to_shift(psize);
#else
if (!is_vm_hugetlb_page(vma))
return PAGE_SIZE;
return huge_page_size(hstate_vma(vma));
#endif
}
static inline bool is_power_of_4(unsigned long x)
{
if (is_power_of_2(x))
return (__ilog2(x) % 2) ? false : true;
return false;
}
static int __init add_huge_page_size(unsigned long long size)
{
int shift = __ffs(size);
int mmu_psize;
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
/* Check that it is a page size supported by the hardware and
* that it fits within pagetable and slice limits. */
#ifdef CONFIG_PPC_FSL_BOOK3E
if ((size < PAGE_SIZE) || !is_power_of_4(size))
return -EINVAL;
#else
if (!is_power_of_2(size)
|| (shift > SLICE_HIGH_SHIFT) || (shift <= PAGE_SHIFT))
return -EINVAL;
#endif
if ((mmu_psize = shift_to_mmu_psize(shift)) < 0)
return -EINVAL;
#ifdef CONFIG_SPU_FS_64K_LS
/* Disable support for 64K huge pages when 64K SPU local store
* support is enabled as the current implementation conflicts.
*/
if (shift == PAGE_SHIFT_64K)
return -EINVAL;
#endif /* CONFIG_SPU_FS_64K_LS */
BUG_ON(mmu_psize_defs[mmu_psize].shift != shift);
/* Return if huge page size has already been setup */
if (size_to_hstate(size))
return 0;
hugetlb_add_hstate(shift - PAGE_SHIFT);
return 0;
}
static int __init hugepage_setup_sz(char *str)
{
unsigned long long size;
size = memparse(str, &str);
if (add_huge_page_size(size) != 0)
printk(KERN_WARNING "Invalid huge page size specified(%llu)\n", size);
return 1;
}
__setup("hugepagesz=", hugepage_setup_sz);
#ifdef CONFIG_PPC_FSL_BOOK3E
struct kmem_cache *hugepte_cache;
static int __init hugetlbpage_init(void)
{
int psize;
for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) {
unsigned shift;
if (!mmu_psize_defs[psize].shift)
continue;
shift = mmu_psize_to_shift(psize);
/* Don't treat normal page sizes as huge... */
if (shift != PAGE_SHIFT)
if (add_huge_page_size(1ULL << shift) < 0)
continue;
}
/*
* Create a kmem cache for hugeptes. The bottom bits in the pte have
* size information encoded in them, so align them to allow this
*/
hugepte_cache = kmem_cache_create("hugepte-cache", sizeof(pte_t),
HUGEPD_SHIFT_MASK + 1, 0, NULL);
if (hugepte_cache == NULL)
panic("%s: Unable to create kmem cache for hugeptes\n",
__func__);
/* Default hpage size = 4M */
if (mmu_psize_defs[MMU_PAGE_4M].shift)
HPAGE_SHIFT = mmu_psize_defs[MMU_PAGE_4M].shift;
else
panic("%s: Unable to set default huge page size\n", __func__);
return 0;
}
#else
static int __init hugetlbpage_init(void)
{
powerpc/mm: Allow more flexible layouts for hugepage pagetables Currently each available hugepage size uses a slightly different pagetable layout: that is, the bottem level table of pointers to hugepages is a different size, and may branch off from the normal page tables at a different level. Every hugepage aware path that needs to walk the pagetables must therefore look up the hugepage size from the slice info first, and work out the correct way to walk the pagetables accordingly. Future hardware is likely to add more possible hugepage sizes, more layout options and more mess. This patch, therefore reworks the handling of hugepage pagetables to reduce this complexity. In the new scheme, instead of having to consult the slice mask, pagetable walking code can check a flag in the PGD/PUD/PMD entries to see where to branch off to hugepage pagetables, and the entry also contains the information (eseentially hugepage shift) necessary to then interpret that table without recourse to the slice mask. This scheme can be extended neatly to handle multiple levels of self-describing "special" hugepage pagetables, although for now we assume only one level exists. This approach means that only the pagetable allocation path needs to know how the pagetables should be set out. All other (hugepage) pagetable walking paths can just interpret the structure as they go. There already was a flag bit in PGD/PUD/PMD entries for hugepage directory pointers, but it was only used for debug. We alter that flag bit to instead be a 0 in the MSB to indicate a hugepage pagetable pointer (normally it would be 1 since the pointer lies in the linear mapping). This means that asm pagetable walking can test for (and punt on) hugepage pointers with the same test that checks for unpopulated page directory entries (beq becomes bge), since hugepage pointers will always be positive, and normal pointers always negative. While we're at it, we get rid of the confusing (and grep defeating) #defining of hugepte_shift to be the same thing as mmu_huge_psizes. Signed-off-by: David Gibson <dwg@au1.ibm.com> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2009-10-26 19:24:31 +00:00
int psize;
if (!mmu_has_feature(MMU_FTR_16M_PAGE))
return -ENODEV;
for (psize = 0; psize < MMU_PAGE_COUNT; ++psize) {
unsigned shift;
unsigned pdshift;
if (!mmu_psize_defs[psize].shift)
continue;
shift = mmu_psize_to_shift(psize);
if (add_huge_page_size(1ULL << shift) < 0)
continue;
if (shift < PMD_SHIFT)
pdshift = PMD_SHIFT;
else if (shift < PUD_SHIFT)
pdshift = PUD_SHIFT;
else
pdshift = PGDIR_SHIFT;
pgtable_cache_add(pdshift - shift, NULL);
if (!PGT_CACHE(pdshift - shift))
panic("hugetlbpage_init(): could not create "
"pgtable cache for %d bit pagesize\n", shift);
}
/* Set default large page size. Currently, we pick 16M or 1M
* depending on what is available
*/
if (mmu_psize_defs[MMU_PAGE_16M].shift)
HPAGE_SHIFT = mmu_psize_defs[MMU_PAGE_16M].shift;
else if (mmu_psize_defs[MMU_PAGE_1M].shift)
HPAGE_SHIFT = mmu_psize_defs[MMU_PAGE_1M].shift;
return 0;
}
#endif
module_init(hugetlbpage_init);
void flush_dcache_icache_hugepage(struct page *page)
{
int i;
void *start;
BUG_ON(!PageCompound(page));
for (i = 0; i < (1UL << compound_order(page)); i++) {
if (!PageHighMem(page)) {
__flush_dcache_icache(page_address(page+i));
} else {
start = kmap_atomic(page+i, KM_PPC_SYNC_ICACHE);
__flush_dcache_icache(start);
kunmap_atomic(start, KM_PPC_SYNC_ICACHE);
}
}
}