Add max CPU nodes field to ibm,client-architecture call
cover Add max CPU nodes field to ibm,client-architecture call
Large NUMA machines require Linux to indicate support for more than 64 cpu cores. This is done through the ibm,client-architecture call, and is documented in the PAPR. There is also another new field added to indicate that the OS is Linux as a hint for possible future performance settings.
The first patch attempts to statically initialize the number of supported cores with NR_CPUS, the maximum Linux could boot if the cpus did not have multi-threading (SMT). It can overestimate by the factor of SMT. For instance on Power6 with 2 way SMT it would overestimate by a factor of 2. The result of this overestimation is that Linux might not be able boot all the cpus assigned to it, but would still boot NR_CPUS worth of SMT threads.
The second patch adjusts for SMT by reading the device tree before unflattening.
I've updated patch 1 from previous comments. Patch 1 can be taken on its own if Patch 2 seems like overkill.
PATCH 1/2 Add static fields to ibm,client-architecture call
This patch adds 2 fields to the ibm_architecture_vec array.
The first of these fields indicates the number of cores which Linux can boot. It does not account for SMT, so it may result in cpus assigned to Linux which cannot be booted. A second patch follows that dynamically updates this for SMT.
The second field just indicates that our OS is Linux, and not another OS. The system may or may not use this hint to performance tune settings for Linux.
arch/powerpc/kernel/prom_init.c +0 -0
@@ -654,6 +656,9 @@ static void __init early_cmdline_parse(v
#define OV5_CMO 0x00
#endif
+/* Option Vector 6: IBM PAPR hints */
+#define OV6_LINUX 0x02 /* Linux is our OS */
+
/*
* The architecture vector has an array of PVR mask/value pairs,
* followed by # option vectors - 1, followed by the option vectors.
@@ -665,7 +670,7 @@ static unsigned char ibm_architecture_ve W(0xffffffff), W(0x0f000003), /* all 2.06-compliant */ W(0xffffffff), W(0x0f000002), /* all 2.05-compliant */ W(0xfffffffe), W(0x0f000001), /* all 2.04-compliant and earlier */
- 5 - 1, /* 5 option vectors */
+ 6 - 1, /* 6 option vectors */
/* option vector 1: processor architectures supported */ 3 - 2, /* length */
@@ -697,12 +702,24 @@ static unsigned char ibm_architecture_ve 0, /* don't halt */
/* option vector 5: PAPR/OF options */
- 5 - 2, /* length */
+ 13 - 2, /* length */
0, /* don't ignore, don't halt */OV5_LPAR | OV5_SPLPAR | OV5_LARGE_PAGES | OV5_DRCONF_MEMORY | OV5_DONATE_DEDICATE_CPU | OV5_MSI,
0,OV5_CMO,
+ 0,
+ 0,
+ 0,
+ 0,
+ W(NR_CPUS), /* number of cores supported*/
+
+ /* option vector 6: IBM PAPR hints */
+ 4 - 2, /* length */
+ 0,
+ 0,
+ OV6_LINUX,
+
};
/* Old method - ELF header with PT_NOTE sections */
PATCH 1/2 Add static fields to ibm,client-architecture call
This patch adds 2 fields to the ibm_architecture_vec array.
The first of these fields indicates the number of cores which Linux can boot. It does not account for SMT, so it may result in cpus assigned to Linux which cannot be booted. A second patch follows that dynamically updates this for SMT.
The second field just indicates that our OS is Linux, and not another OS. The system may or may not use this hint to performance tune settings for Linux.
Sorry if the last version sent out got whitespace munged. Hopefully evolution works on this one.
arch/powerpc/kernel/prom_init.c +17 -2
@@ -654,6 +656,9 @@ static void __init early_cmdline_parse(v#define OV5_CMO 0x00#endif +/* Option Vector 6: IBM PAPR hints */ +#define OV6_LINUX 0x02/* Linux is our OS */ +/* * The architecture vector has an array of PVR mask/value pairs, * followed by # option vectors - 1, followed by the option vectors.
@@ -665,7 +670,7 @@ static unsigned char ibm_architecture_veW (0xffffffff ),W (0x0f000003 ),/* all 2.06-compliant */ W (0xffffffff ),W (0x0f000002 ),/* all 2.05-compliant */ W (0xfffffffe ),W (0x0f000001 ),/* all 2.04-compliant and earlier */ -5 -1 ,/* 5 option vectors */ +6 -1 ,/* 6 option vectors */ /* option vector 1: processor architectures supported */ 3 - 2, /* length */
@@ -697,12 +702,24 @@ static unsigned char ibm_architecture_ve0 ,/* don't halt */ /* option vector 5: PAPR/OF options */ -5 -2 ,/* length */ +13 -2 ,/* length */ 0 ,/* don't ignore, don't halt */ OV5_LPAR |OV5_SPLPAR |OV5_LARGE_PAGES |OV5_DRCONF_MEMORY |OV5_DONATE_DEDICATE_CPU |OV5_MSI ,0 ,OV5_CMO , +0 , +0 , +0 , +0 , +W (NR_CPUS ),/* number of cores supported*/ + +/* option vector 6: IBM PAPR hints */ +4 -2 ,/* length */ +0 , +0 , + OV6_LINUX, + };/* Old method - ELF header with PT_NOTE sections */
Re: [PATCH 1/2] Add static fields to ibm,client-architecture call
Quoted 10 lines · Joel Schopp
This patch adds 2 fields to the
ibm_architecture_vecarray.The first of these fields indicates the number of cores which Linux can boot. It does not account for SMT, so it may result in cpus assigned to Linux which cannot be booted. A second patch follows that dynamically updates this for SMT.
The second field just indicates that our OS is Linux, and not another OS. The system may or may not use this hint to performance tune settings for Linux.
Patch submission 101...
Quoted 2 lines
Sorry if the last version sent out got whitespace munged. Hopefully evolution works on this one.
The above should be past the signature marker (--) so it doesn't end up in the changeset comments
Quoted 1 line
Signed-off-by Joel Schopp<[email protected]>
The marker should be here or the line below:
Quoted 2 lines
Index:
linux-2.6.git/arch/powerpc/kernel/prom_init.c
===================================================================
Would also end up in the comments.
The rest looks fine.
Cheers,
Ben.
Quoted 53 lines
arch/powerpc/kernel/prom_init.c +17
-2@@ -654,6 +656,9 @@ static void __init early_cmdline_parse(v#define OV5_CMO 0x00#endif +/* Option Vector 6: IBM PAPR hints */ +#define OV6_LINUX 0x02/* Linux is our OS */ +/* * The architecture vector has an array of PVR mask/value pairs, * followed by # option vectors - 1, followed by the option vectors. @@ -665,7 +670,7 @@ static unsigned char ibm_architecture_veW (0xffffffff ),W (0x0f000003 ),/* all 2.06-compliant */ W (0xffffffff ),W (0x0f000002 ),/* all 2.05-compliant */ W (0xfffffffe ),W (0x0f000001 ),/* all 2.04-compliant and earlier */ -5 -1 ,/* 5 option vectors */ +6 -1 ,/* 6 option vectors */ /* option vector 1: processor architectures supported */ 3 - 2, /* length */@@ -697,12 +702,24 @@ static unsigned char ibm_architecture_ve0 ,/* don't halt */ /* option vector 5: PAPR/OF options */ -5 -2 ,/* length */ +13 -2 ,/* length */ 0 ,/* don't ignore, don't halt */ OV5_LPAR |OV5_SPLPAR |OV5_LARGE_PAGES |OV5_DRCONF_MEMORY |OV5_DONATE_DEDICATE_CPU |OV5_MSI ,0 ,OV5_CMO , +0 , +0 , +0 , +0 , +W (NR_CPUS ),/* number of cores supported*/ + +/* option vector 6: IBM PAPR hints */ +4 -2 ,/* length */ +0 , +0 , + OV6_LINUX, + };/* Old method - ELF header with PT_NOTE sections */ _______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev
Re: [PATCH 1/2] Add static fields to ibm,client-architecture call
Oh and :
Quoted 1 line
Quoted 1 line
Signed-off-by Joel Schopp<[email protected]>
^ space ^ space
Thanks !
Cheers,
Ben.
Quoted 66 lines
The marker should be here or the line below:
Quoted 2 lines
Index:
linux-2.6.git/arch/powerpc/kernel/prom_init.c
===================================================================Would also end up in the comments.
The rest looks fine.
Cheers,
Ben.Quoted 53 lines
arch/powerpc/kernel/prom_init.c +17
-2@@ -654,6 +656,9 @@ static void __init early_cmdline_parse(v#define OV5_CMO 0x00#endif +/* Option Vector 6: IBM PAPR hints */ +#define OV6_LINUX 0x02/* Linux is our OS */ +/* * The architecture vector has an array of PVR mask/value pairs, * followed by # option vectors - 1, followed by the option vectors. @@ -665,7 +670,7 @@ static unsigned char ibm_architecture_veW (0xffffffff ),W (0x0f000003 ),/* all 2.06-compliant */ W (0xffffffff ),W (0x0f000002 ),/* all 2.05-compliant */ W (0xfffffffe ),W (0x0f000001 ),/* all 2.04-compliant and earlier */ -5 -1 ,/* 5 option vectors */ +6 -1 ,/* 6 option vectors */ /* option vector 1: processor architectures supported */ 3 - 2, /* length */@@ -697,12 +702,24 @@ static unsigned char ibm_architecture_ve0 ,/* don't halt */ /* option vector 5: PAPR/OF options */ -5 -2 ,/* length */ +13 -2 ,/* length */ 0 ,/* don't ignore, don't halt */ OV5_LPAR |OV5_SPLPAR |OV5_LARGE_PAGES |OV5_DRCONF_MEMORY |OV5_DONATE_DEDICATE_CPU |OV5_MSI ,0 ,OV5_CMO , +0 , +0 , +0 , +0 , +W (NR_CPUS ),/* number of cores supported*/ + +/* option vector 6: IBM PAPR hints */ +4 -2 ,/* length */ +0 , +0 , + OV6_LINUX, + };/* Old method - ELF header with PT_NOTE sections */ _______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev
PATCH v2 1/2 Add static fields to ibm,client-architecture call
In reply to Joel SchoppThis patch adds 2 fields to the ibm_architecture_vec array.
The first of these fields indicates the number of cores which Linux can boot. It does not account for SMT, so it may result in cpus assigned to Linux which cannot be booted. A second patch follows that dynamically updates this for SMT.
The second field just indicates that our OS is Linux, and not another OS. The system may or may not use this hint to performance tune settings for Linux.
---
arch/powerpc/kernel/prom_init.c +17 -2
@@ -654,6 +656,9 @@ static void __init early_cmdline_parse(v#define OV5_CMO 0x00#endif +/* Option Vector 6: IBM PAPR hints */ +#define OV6_LINUX 0x02/* Linux is our OS */ +/* * The architecture vector has an array of PVR mask/value pairs, * followed by # option vectors - 1, followed by the option vectors.
@@ -665,7 +670,7 @@ static unsigned char ibm_architecture_veW (0xffffffff ),W (0x0f000003 ),/* all 2.06-compliant */ W (0xffffffff ),W (0x0f000002 ),/* all 2.05-compliant */ W (0xfffffffe ),W (0x0f000001 ),/* all 2.04-compliant and earlier */ -5 -1 ,/* 5 option vectors */ +6 -1 ,/* 6 option vectors */ /* option vector 1: processor architectures supported */ 3 - 2, /* length */
@@ -697,12 +702,24 @@ static unsigned char ibm_architecture_ve0 ,/* don't halt */ /* option vector 5: PAPR/OF options */ -5 -2 ,/* length */ +13 -2 ,/* length */ 0 ,/* don't ignore, don't halt */ OV5_LPAR |OV5_SPLPAR |OV5_LARGE_PAGES |OV5_DRCONF_MEMORY |OV5_DONATE_DEDICATE_CPU |OV5_MSI ,0 ,OV5_CMO , +0 , +0 , +0 , +0 , +W (NR_CPUS ),/* number of cores supported*/ + +/* option vector 6: IBM PAPR hints */ +4 -2 ,/* length */ +0 , +0 , + OV6_LINUX, + };/* Old method - ELF header with PT_NOTE sections */
PATCH 2/2 Update ibm,client-architecture call field based on device tree
In reply to Joel SchoppIn the previous patch the client-architecture field for the number of cores supported is set statically as high as is possible. However, that static setting could be too high if the system supports smt, resulting in cpus assigned to Linux that are not booted. This patch reads the device tree (before it is unflattened) to determine the amount of smt. It then dynamically updates the entires in the array with the proper number of cores supported. Tests show this correctly detecting SMT4 on a Power7 and still booting all the supported cores on a large machine.
arch/powerpc/kernel/prom_init.c +49 -0
@@ -141,6 +141,8 @@ typedef u32 cell_t;extern void __start (unsigned long r3,unsigned long r4,unsigned long r5); +static int __init prom_smt_way (void ); +#ifdef CONFIG_PPC64 extern int enter_prom (struct prom_args *args,unsigned long entry);#else
@@ -811,9 +813,17 @@ static void __init prom_send_capabilitie {ihandle elfloader, root;prom_arg_t ret; +u32 *cores; root =call_prom ("open" ,1 ,1 ,ADDR ("/" ));if (root !=0 ) { +/* +* If you add to the struct, please be sure the 100 index +* didn't change. The BUILD_BUG_ON is a reminder. +*/ +BUILD_BUG_ON (sizeof (ibm_architecture_vec) !=108 ); + cores = (u32 *) &ibm_architecture_vec[100 ]; + *cores = (u32) (NR_CPUS /prom_smt_way ());/* try calling the ibm,client-architecture-support method */ prom_printf ("Calling ibm,client-architecture-support..." ); if (call_prom_ret("call-method", 3, 2, &ret,
@@ -1031,6 +1041,45 @@ static void __init reserve_mem(u64 base,RELOC (mem_reserve_cnt) = cnt +1 ; } + +static int __init prom_smt_way (void ) +{ +phandle node; +char type[64 ]; +unsigned int plen; + +for (node =0 ;prom_next_node (&node); ) { + type[0 ] =0 ; +prom_getprop (node,"device_type" , type,sizeof (type)); + +if (type[0 ] ==0 ) { +/* +* CHRP Longtrail machines have no device_type +* on the memory node, so check the name instead... +*/ +prom_getprop (node,"name" , type,sizeof (type)); + } +if (strcmp (type,RELOC ("cpu" ))) +continue ; + +/* +* There is an entry for each smt thread, each entry being +* 4 bytes long. All cpus should have the same number of +* smt threads, so return after finding the first. +*/ + plen =prom_getproplen (node,"ibm,ppc-interrupt-server#s" ); +prom_debug ("smt %x\n" , (unsigned long ) plen); +if (plen >=4 ) +return plen /4 ; + } +/* +* If things go wrong and we get here fallback to SMT1 +*/ +prom_debug ("unable to determine smt from device tree, guessing smt1\n" ); +return 1 ; + +} +/* * Initialize memory allocation mechanism, parse "memory" nodes and * obtain that way the top of memory and RMO to setup out local allocator
Re: [PATCH 2/2] Update ibm,client-architecture call field based on device tree
Quoted 8 lines · Joel Schopp
In the previous patch the client-architecture field for the number of cores supported is set statically as high as is possible. However, that static setting could be too high if the system supports smt, resulting in cpus assigned to Linux that are not booted. This patch reads the device tree (before it is unflattened) to determine the amount of smt. It then dynamically updates the entires in the array with the proper number of cores supported. Tests show this correctly detecting SMT4 on a Power7 and still booting all the supported cores on a large machine.
Same comments about submission format as the previous patch...
arch/powerpc/kernel/prom_init.c typedef u32 cell_t; · line 144
Quoted 10 lines
Signed-off-by Joel Schopp<[email protected]>Index:
linux-2.6.git/arch/powerpc/kernel/prom_init.c
===================================================================arch/powerpc/kernel/prom_init.c +1
-0@@ -141,6 +141,8 @@ typedef u32 cell_t;extern void __start (unsigned long r3,unsigned long r4,unsigned long r5); +static int __init prom_smt_way (void );
Do you really need a forward declaration ? We generally avoid those and prefer instead having the functions in the right order to make it unnecessary.
Quoted 6 lines
arch/powerpc/kernel/prom_init.c +0
-0#ifdef CONFIG_PPC64 extern int enter_prom (struct prom_args *args,unsigned long entry);#else @@ -811,9 +813,17 @@ static void __init prom_send_capabilitie {ihandle elfloader, root;
arch/powerpc/kernel/prom_init.c static void __init prom_send_capabilitie · line 824
Quoted 10 lines
arch/powerpc/kernel/prom_init.c +6
-0prom_arg_t ret; +u32 *cores; root =call_prom ("open" ,1 ,1 ,ADDR ("/" ));if (root !=0 ) { +/* +* If you add to the struct, please be sure the 100 index +* didn't change. The BUILD_BUG_ON is a reminder. +*/ +BUILD_BUG_ON (sizeof (ibm_architecture_vec) !=108 );
This is indeed a bit fishy... a nicer way may have been to have the vector in an asm file with labels but that's probably overkill. Just maybe add a runtime test that checks the value read initially is NR_CORES and if not, print a big fat warning ?
arch/powerpc/kernel/prom_init.c static void __init prom_send_capabilitie · line 826
Quoted 2 lines
arch/powerpc/kernel/prom_init.c +2
-0+ cores = (u32 *) &ibm_architecture_vec[100 ]; + *cores = (u32) (NR_CPUS /prom_smt_way ());
The style is a bit gross though I suppose it will do. Drop the cast on the second line, it's not useful, and stick some spaces around that division. Also maybe print out a message saying to what value you adjusted the max number of supported cores, might be useful on the field to diagnose issues.
Quoted 14 lines
arch/powerpc/kernel/prom_init.c +7
-0/* try calling the ibm,client-architecture-support method */ prom_printf ("Calling ibm,client-architecture-support..." ); if (call_prom_ret("call-method", 3, 2, &ret,@@ -1031,6 +1041,45 @@ static void __init reserve_mem(u64 base,RELOC (mem_reserve_cnt) = cnt +1 ; } + +static int __init prom_smt_way (void ) +{ +phandle node; +char type[64 ]; +unsigned int plen; +
arch/powerpc/kernel/prom_init.c static void __init reserve_mem(u64 base, · line 1060
Quoted 10 lines
arch/powerpc/kernel/prom_init.c +10
-0+for (node =0 ;prom_next_node (&node); ) { + type[0 ] =0 ; +prom_getprop (node,"device_type" , type,sizeof (type)); + +if (type[0 ] ==0 ) { +/* +* CHRP Longtrail machines have no device_type +* on the memory node, so check the name instead... +*/ +prom_getprop (node,"name" , type,sizeof (type));
They also have no SMT :-) Just continue instead. You could also have instead gone for /cpus and peeked at the first child. Might have lead to a simpler construct (and faster runtime)
arch/powerpc/kernel/prom_init.c static void __init reserve_mem(u64 base, · line 1071
Quoted 11 lines
arch/powerpc/kernel/prom_init.c +11
-0+ } +if (strcmp (type,RELOC ("cpu" ))) +continue ; + +/* +* There is an entry for each smt thread, each entry being +* 4 bytes long. All cpus should have the same number of +* smt threads, so return after finding the first. +*/ + plen =prom_getproplen (node,"ibm,ppc-interrupt-server#s" ); +prom_debug ("smt %x\n" , (unsigned long ) plen);
Might only be a debug message but it should be a tad more verbose about what it's actually printing.
arch/powerpc/kernel/prom_init.c static void __init reserve_mem(u64 base, · line 1082
Quoted 11 lines
arch/powerpc/kernel/prom_init.c +11
-0+if (plen >=4 ) +return plen /4 ; + } +/* +* If things go wrong and we get here fallback to SMT1 +*/ +prom_debug ("unable to determine smt from device tree, guessing smt1\n" ); +return 1 ; + +} +
Cheers,
Ben.
Quoted 9 lines
arch/powerpc/kernel/prom_init.c +0
-0/* * Initialize memory allocation mechanism, parse "memory" nodes and * obtain that way the top of memory and RMO to setup out local allocator _______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev
PATCH v2 2/2 Update ibm,client-architecture call field based on device tree
In reply to Joel SchoppIn the previous patch the client-architecture field for the number of cores supported is set statically as high as is possible. However, that static setting could be too high if the system supports smt, resulting in cpus assigned to Linux that are not booted. This patch reads the device tree (before it is unflattened) to determine the amount of smt. It then dynamically updates the entires in the array with the proper number of cores supported. Tests show this correctly detecting SMT4 on a Power7 and still booting all the supported cores on a large machine.
---
arch/powerpc/kernel/prom_init.c +42 -0
@@ -807,13 +807,55 @@ static struct fake_elf { } }; +static int __init prom_smt_way (void ) +{ +phandle node; +char type[64 ]; +unsigned int plen; + +for (node =0 ;prom_next_node (&node); ) { + type[0 ] =0 ; +prom_getprop (node,"device_type" , type,sizeof (type)); + +if (strcmp (type,RELOC ("cpu" ))) +continue ; + +/* +* There is an entry for each smt thread, each entry being +* 4 bytes long. All cpus should have the same number of +* smt threads, so return after finding the first. +*/ + plen =prom_getproplen (node,"ibm,ppc-interrupt-server#s" ); +prom_printf ("smt threads per core %x\n" , (unsigned long ) plen); +if (plen >=4 ) +return plen /4 ; + } +/* +* If things go wrong and we get here fallback to SMT1 +*/ +prom_debug ("unable to determine smt from device tree, guessing smt1\n" ); +return 1 ; + +} + +static void __init prom_send_capabilities (void ) {ihandle elfloader, root;prom_arg_t ret; +u32 *cores; root =call_prom ("open" ,1 ,1 ,ADDR ("/" ));if (root !=0 ) { +/* +* If you add to the struct, please be sure the 100 index +* didn't change. The BUILD_BUG_ON is a reminder. +*/ + cores = (u32 *) &ibm_architecture_vec[100 ]; +if (*cores !=NR_CPUS ) +prom_printf ("client-architecture structure corrupted\n" ); + *cores = (NR_CPUS /prom_smt_way ()); +prom_printf ("setting client-architecture cores to %x\n" , *cores);/* try calling the ibm,client-architecture-support method */ prom_printf ("Calling ibm,client-architecture-support..." ); if (call_prom_ret("call-method", 3, 2, &ret,
Re: [PATCHv2 2/2] Update ibm,client-architecture call field based on device tree
Quoted 7 lines · Joel Schopp
static void
__initprom_send_capabilities(void)
{arch/powerpc/kernel/prom_init.c +1
-0ihandle elfloader, root;prom_arg_t ret; +u32 *cores; root =call_prom ("open" ,1 ,1 ,ADDR ("/" ));
arch/powerpc/kernel/prom_init.c static struct fake_elf { · line 858
Quoted 10 lines
arch/powerpc/kernel/prom_init.c +9
-0if (root !=0 ) { +/* +* If you add to the struct, please be sure the 100 index +* didn't change. The BUILD_BUG_ON is a reminder. +*/ + cores = (u32 *) &ibm_architecture_vec[100 ]; +if (*cores !=NR_CPUS ) +prom_printf ("client-architecture structure corrupted\n" ); + *cores = (NR_CPUS /prom_smt_way ()); +prom_printf ("setting client-architecture cores to %x\n" , *cores);
I don't know if I'm painting a bike shed of if this is a real concern, but if *cores isn't NR_CPUS shouldn't we do nothing rather then clobbering it?
Yours Tony
Re: [PATCHv2 2/2] Update ibm,client-architecture call field based on device tree
Quoted 10 lines
arch/powerpc/kernel/prom_init.c static struct fake_elf { · line 858
Quoted 6 lines
arch/powerpc/kernel/prom_init.c +4
-0+if (*cores !=NR_CPUS ) +prom_printf ("client-architecture structure corrupted\n" ); + *cores = (NR_CPUS /prom_smt_way ()); +prom_printf ("setting client-architecture cores to %x\n" , *cores);I don't know if I'm painting a bike shed of if this is a real concern, but if *cores isn't
NR_CPUSshouldn't we do nothing rather then clobbering it?
Quoted 2 lines
Yours Tony
If it isn't NR_CPUS we're pretty broken if we set it or if we don't. My previous version did a BUILD_BUG_ON() but Ben didn't like that and said he preferred just a warning message.
Re: [PATCHv2 2/2] Update ibm,client-architecture call field based on device tree
Quoted 9 lines · Joel Schopp
Quoted 9 lines
arch/powerpc/kernel/prom_init.c static struct fake_elf { · line 858
Quoted 5 lines
arch/powerpc/kernel/prom_init.c +4
-0+if (*cores !=NR_CPUS ) +prom_printf ("client-architecture structure corrupted\n" ); + *cores = (NR_CPUS /prom_smt_way ()); +prom_printf ("setting client-architecture cores to %x\n" , *cores);I don't know if I'm painting a bike shed of if this is a real concern, but if *cores isn't
NR_CPUSshouldn't we do nothing rather then clobbering it?
Quoted 5 lines
Quoted 2 lines
Yours Tony
If it isn't
NR_CPUSwe're pretty broken if we set it or if we don't. My previous version did aBUILD_BUG_ON()but Ben didn't like that and said he preferred just a warning message.
BUILD_BUG_ON would probably not have worked unless gcc smarter than I think it is :-) Your latest approach is bad because when you detect you are poking at the wrong thing ... you still proceed and do it. In fact you -are- poking at the wrong thing always since you are missing PTRRELOC() to relocate the pointer to the vec.
I'll post a new variant of that patch.
Cheers,
Ben.
v2 cover Add max CPU nodes field to ibm,client-architecture call
In reply to Joel SchoppLarge NUMA machines require Linux to indicate support for more than 64 cpu cores. This is done through the ibm,client-architecture call, and is documented in the PAPR. There is also another new field added to indicate that the OS is Linux as a hint for possible future performance settings.
The first patch attempts to statically initialize the number of supported cores with NR_CPUS, the maximum Linux could boot if the cpus did not have multi-threading (SMT). It can overestimate by the factor of SMT. For instance on Power6 with 2 way SMT it would overestimate by a factor of 2. The result of this overestimation is that Linux might not be able boot all the cpus assigned to it, but would still boot NR_CPUS worth of SMT threads.
The second patch adjusts for SMT by reading the device tree before unflattening.
I've updated patch 1 from previous comments. Patch 1 can be taken on its own if Patch 2 seems like overkill.
v2 - Updated patch 2 style and messages. Basic functionality is the same as v1.