| Message ID | 1341839621-28332-12-git-send-email-aneesh.kumar@linux.vnet.ibm.com (mailing list archive) |
|---|---|
| State | Changes Requested |
| Delegated to: | Benjamin Herrenschmidt |
| Headers | show |
On Mon, Jul 09, 2012 at 06:43:41PM +0530, Aneesh Kumar K.V wrote: > From: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com> > > Increase max addressable range to 64TB. This is not tested on > real hardware yet. > > Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> > --- > arch/powerpc/include/asm/mmu-hash64.h | 8 ++++---- > arch/powerpc/include/asm/pgtable-ppc64-4k.h | 2 +- > arch/powerpc/include/asm/pgtable-ppc64-64k.h | 2 +- > arch/powerpc/include/asm/processor.h | 4 ++-- > arch/powerpc/include/asm/sparsemem.h | 4 ++-- > 5 files changed, 10 insertions(+), 10 deletions(-) > > diff --git a/arch/powerpc/include/asm/mmu-hash64.h b/arch/powerpc/include/asm/mmu-hash64.h > index aa0d560..a227ba7 100644 > --- a/arch/powerpc/include/asm/mmu-hash64.h > +++ b/arch/powerpc/include/asm/mmu-hash64.h > @@ -374,16 +374,16 @@ extern void slb_set_size(u16 size); > */ > > #define VSID_MULTIPLIER_256M ASM_CONST(200730139) /* 28-bit prime */ > -#define VSID_BITS_256M 36 > +#define VSID_BITS_256M 38 > #define VSID_MODULUS_256M ((1UL<<VSID_BITS_256M)-1) With these settings, the multiplication in ASM_VSID_SCRAMBLE could overflow, leading to incorrect results (which would cause occasional corruption of user processes under heavy load). You will need to reduce the multiplier to be less than 2^26, and it will need to be co-prime with 2^38 - 1. (Probably, the same value as we use in the 1T case would be OK.) Paul.
Paul Mackerras <paulus@samba.org> writes: > On Mon, Jul 09, 2012 at 06:43:41PM +0530, Aneesh Kumar K.V wrote: >> From: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com> >> >> Increase max addressable range to 64TB. This is not tested on >> real hardware yet. >> >> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> >> --- >> arch/powerpc/include/asm/mmu-hash64.h | 8 ++++---- >> arch/powerpc/include/asm/pgtable-ppc64-4k.h | 2 +- >> arch/powerpc/include/asm/pgtable-ppc64-64k.h | 2 +- >> arch/powerpc/include/asm/processor.h | 4 ++-- >> arch/powerpc/include/asm/sparsemem.h | 4 ++-- >> 5 files changed, 10 insertions(+), 10 deletions(-) >> >> diff --git a/arch/powerpc/include/asm/mmu-hash64.h b/arch/powerpc/include/asm/mmu-hash64.h >> index aa0d560..a227ba7 100644 >> --- a/arch/powerpc/include/asm/mmu-hash64.h >> +++ b/arch/powerpc/include/asm/mmu-hash64.h >> @@ -374,16 +374,16 @@ extern void slb_set_size(u16 size); >> */ >> >> #define VSID_MULTIPLIER_256M ASM_CONST(200730139) /* 28-bit prime */ >> -#define VSID_BITS_256M 36 >> +#define VSID_BITS_256M 38 >> #define VSID_MODULUS_256M ((1UL<<VSID_BITS_256M)-1) > > With these settings, the multiplication in ASM_VSID_SCRAMBLE could > overflow, leading to incorrect results (which would cause occasional > corruption of user processes under heavy load). You will need to > reduce the multiplier to be less than 2^26, and it will need to be > co-prime with 2^38 - 1. (Probably, the same value as we use in the 1T > case would be OK.) I ended up using the same value as VSID_MULTIPLIER_1T. (gdb) p/d 1ull << 38 $1 = 274877906944 (gdb) p/d 274877906943/12538073 $2 = 21923 (gdb) p/d 12538073ll*21923 $5 = 274872174379
On Mon, Jul 09, 2012 at 06:43:41PM +0530, Aneesh Kumar K.V wrote: > -#define USER_ESID_BITS 16 > -#define USER_ESID_BITS_1T 4 > +#define USER_ESID_BITS 18 > +#define USER_ESID_BITS_1T 6 You also need to change the proto-VSID generation for kernel addresses when you do this. If you don't you'll end up with some user processes using the same VSIDs as we use for the kernel addresses, meaning that those processes won't run very well... Paul.
Paul Mackerras <paulus@samba.org> writes: > On Mon, Jul 09, 2012 at 06:43:41PM +0530, Aneesh Kumar K.V wrote: > >> -#define USER_ESID_BITS 16 >> -#define USER_ESID_BITS_1T 4 >> +#define USER_ESID_BITS 18 >> +#define USER_ESID_BITS_1T 6 > > You also need to change the proto-VSID generation for kernel addresses > when you do this. If you don't you'll end up with some user processes > using the same VSIDs as we use for the kernel addresses, meaning that > those processes won't run very well... > Can you explain this more. right now we generate vsid as below vsid_scramble(ea >> SID_SHIFT, 256M) for kernel vsid_scramble((context << USER_ESID_BITS) | (ea >> SID_SHIFT), 256M); for user what changes are you suggesting ? -aneesh
On Mon, Jul 23, 2012 at 03:52:05PM +0530, Aneesh Kumar K.V wrote: > Paul Mackerras <paulus@samba.org> writes: > > > On Mon, Jul 09, 2012 at 06:43:41PM +0530, Aneesh Kumar K.V wrote: > > > >> -#define USER_ESID_BITS 16 > >> -#define USER_ESID_BITS_1T 4 > >> +#define USER_ESID_BITS 18 > >> +#define USER_ESID_BITS_1T 6 > > > > You also need to change the proto-VSID generation for kernel addresses > > when you do this. If you don't you'll end up with some user processes > > using the same VSIDs as we use for the kernel addresses, meaning that > > those processes won't run very well... > > > > Can you explain this more. right now we generate vsid as below > > vsid_scramble(ea >> SID_SHIFT, 256M) for kernel > > vsid_scramble((context << USER_ESID_BITS) | (ea >> SID_SHIFT), 256M); > for user > > what changes are you suggesting ? Think about it. With the current values of USER_ESID_BITS and CONTEXT_BITS, and the addresses we use for kernel mappings, there are no values of context, user_ea and kernel_ea for which kernel_ea >> SID_SHIFT == (context << USER_ESID_BITS) | (user_ea >> SID_SHIFT) If you increase USER_ESID_BITS, then there will be some context values for which that equation becomes true. For example, if you increase USER_ESID_BITS to 18, then context 0x30000 will generate the same proto-VSIDs as the kernel linear mapping. Since we can hand out contexts up to 0x7ffff (with CONTEXT_BITS = 19), there is a collision. In other words, the proto-VSID space (the space of values that are input to vsid_scramble) is currently divided into two mutually exclusive regions: from 0 to 2^35 - 1 for user processes, and from 2^35 to 2^36 - 1 for kernel addresses. You are wanting to expand the amount of proto-VSID space that user processes can use, but you need either to move the kernel portion of the space, or to make sure that the context allocator doesn't hand out context values that would collide with the kernel portion of the space (or both). Paul.
diff --git a/arch/powerpc/include/asm/mmu-hash64.h b/arch/powerpc/include/asm/mmu-hash64.h index aa0d560..a227ba7 100644 --- a/arch/powerpc/include/asm/mmu-hash64.h +++ b/arch/powerpc/include/asm/mmu-hash64.h @@ -374,16 +374,16 @@ extern void slb_set_size(u16 size); */ #define VSID_MULTIPLIER_256M ASM_CONST(200730139) /* 28-bit prime */ -#define VSID_BITS_256M 36 +#define VSID_BITS_256M 38 #define VSID_MODULUS_256M ((1UL<<VSID_BITS_256M)-1) #define VSID_MULTIPLIER_1T ASM_CONST(12538073) /* 24-bit prime */ -#define VSID_BITS_1T 24 +#define VSID_BITS_1T 26 #define VSID_MODULUS_1T ((1UL<<VSID_BITS_1T)-1) #define CONTEXT_BITS 19 -#define USER_ESID_BITS 16 -#define USER_ESID_BITS_1T 4 +#define USER_ESID_BITS 18 +#define USER_ESID_BITS_1T 6 #define USER_VSID_RANGE (1UL << (USER_ESID_BITS + SID_SHIFT)) diff --git a/arch/powerpc/include/asm/pgtable-ppc64-4k.h b/arch/powerpc/include/asm/pgtable-ppc64-4k.h index 6eefdcf..b3eccf2 100644 --- a/arch/powerpc/include/asm/pgtable-ppc64-4k.h +++ b/arch/powerpc/include/asm/pgtable-ppc64-4k.h @@ -7,7 +7,7 @@ */ #define PTE_INDEX_SIZE 9 #define PMD_INDEX_SIZE 7 -#define PUD_INDEX_SIZE 7 +#define PUD_INDEX_SIZE 9 #define PGD_INDEX_SIZE 9 #ifndef __ASSEMBLY__ diff --git a/arch/powerpc/include/asm/pgtable-ppc64-64k.h b/arch/powerpc/include/asm/pgtable-ppc64-64k.h index 90533dd..be4e287 100644 --- a/arch/powerpc/include/asm/pgtable-ppc64-64k.h +++ b/arch/powerpc/include/asm/pgtable-ppc64-64k.h @@ -7,7 +7,7 @@ #define PTE_INDEX_SIZE 12 #define PMD_INDEX_SIZE 12 #define PUD_INDEX_SIZE 0 -#define PGD_INDEX_SIZE 4 +#define PGD_INDEX_SIZE 6 #ifndef __ASSEMBLY__ #define PTE_TABLE_SIZE (sizeof(real_pte_t) << PTE_INDEX_SIZE) diff --git a/arch/powerpc/include/asm/processor.h b/arch/powerpc/include/asm/processor.h index 413a5ea..ac3861b 100644 --- a/arch/powerpc/include/asm/processor.h +++ b/arch/powerpc/include/asm/processor.h @@ -97,8 +97,8 @@ extern struct task_struct *last_task_used_spe; #endif #ifdef CONFIG_PPC64 -/* 64-bit user address space is 44-bits (16TB user VM) */ -#define TASK_SIZE_USER64 (0x0000100000000000UL) +/* 64-bit user address space is 46-bits (64TB user VM) */ +#define TASK_SIZE_USER64 (0x0000400000000000UL) /* * 32-bit user address space is 4GB - 1 page diff --git a/arch/powerpc/include/asm/sparsemem.h b/arch/powerpc/include/asm/sparsemem.h index 0c5fa31..f6fc0ee 100644 --- a/arch/powerpc/include/asm/sparsemem.h +++ b/arch/powerpc/include/asm/sparsemem.h @@ -10,8 +10,8 @@ */ #define SECTION_SIZE_BITS 24 -#define MAX_PHYSADDR_BITS 44 -#define MAX_PHYSMEM_BITS 44 +#define MAX_PHYSADDR_BITS 46 +#define MAX_PHYSMEM_BITS 46 #endif /* CONFIG_SPARSEMEM */