RISC-V: Show CPU ID and Hart ID separately in /proc/cpuinfo
PATCH v2 1/1 RISC-V: Show CPU ID and Hart ID separately in /proc/cpuinfo
Currently, /proc/cpuinfo show logical CPU ID as Hart ID which is in-correct. This patch shows CPU ID and Hart ID separately in /proc/cpuinfo using cpuid_to_hardid_map().
With this patch, contents of /proc/cpuinfo looks as follows:
processor : 0
hart : 1
isa : rv64imafdc
mmu : sv48
processor : 1
hart : 0
isa : rv64imafdc
mmu : sv48
processor : 2
hart : 2
isa : rv64imafdc
mmu : sv48
processor : 3
hart : 3
isa : rv64imafdc
mmu : sv48
---
Changes since v1:
- Show logical CPU ID as "processor" attribute in /proc/cpuinfo
arch/riscv/kernel/cpu.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)
arch/riscv/kernel/cpu.c +6 -4
@@ -81,7 +81,7 @@ static void print_isa(struct seq_file *f, const char *orig_isa)#endif /* Print the base ISA, as we already know it's legal. */ -seq_puts (f,"isa\t: " ); +seq_puts (f,"isa\t\t: " );seq_write (f, isa,5 ); isa +=5 ;
@@ -96,6 +96,7 @@ static void print_isa(struct seq_file *f, const char *orig_isa) isa++; } } +seq_puts (f,"\n" );/* * If we were given an unsupported ISA in the device tree then print
@@ -116,7 +117,7 @@ static void print_mmu(struct seq_file *f, const char *mmu_type)return ;#endif -seq_printf (f,"mmu\t: %s\n" , mmu_type+6 ); +seq_printf (f,"mmu\t\t: %s\n" , mmu_type+6 ); }static void *c_start (struct seq_file *m,loff_t *pos)
@@ -144,14 +145,15 @@ static int c_show(struct seq_file *m, void *v) NULL);const char *compat, *isa, *mmu; -seq_printf (m,"hart\t: %lu\n" , cpu_id); +seq_printf (m,"processor\t: %lu\n" , cpu_id); +seq_printf (m,"hart\t\t: %lu\n" ,cpuid_to_hardid_map (cpu_id));if (!of_property_read_string (node,"riscv,isa" , &isa))print_isa (m, isa);if (!of_property_read_string (node,"mmu-type" , &mmu))print_mmu (m, mmu);if (!of_property_read_string (node,"compatible" , &compat) &&strcmp (compat,"riscv" )) -seq_printf (m,"uarch\t: %s\n" , compat); +seq_printf (m,"uarch\t\t: %s\n" , compat);seq_puts (m,"\n" );return 0 ;
Signature · 1 line
2.17.1
Quoted 4 lines · Anup Patel
Currently,
/proc/cpuinfoshow logical CPU ID as Hart ID which is in-correct. This patch shows CPU ID and Hart ID separately in/proc/cpuinfousingcpuid_to_hardid_map().
I noticed it should be cpuid_to_hartid_map instead of cpuid_to_hardid_map. It was a typo in my smp cleanup patch series. Sorry for the inconvenience here.
I can include this patch in into my series fixing the typo if you want.
Regards,
Atish
Quoted 79 lines
With this patch, contents of
/proc/cpuinfolooks as follows:
processor : 0
hart : 1
isa : rv64imafdc
mmu : sv48processor : 1
hart : 0
isa : rv64imafdc
mmu : sv48processor : 2
hart : 2
isa : rv64imafdc
mmu : sv48processor : 3
hart : 3
isa : rv64imafdc
mmu : sv48Signed-off-by Anup Patel <[email protected]>---
Changes since v1:
- Show logical CPU ID as "processor" attribute in/proc/cpuinfoarch/riscv/kernel/cpu.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)arch/riscv/kernel/cpu.c +6
-4@@ -81,7 +81,7 @@ static void print_isa(struct seq_file *f, const char *orig_isa)#endif /* Print the base ISA, as we already know it's legal. */ -seq_puts (f,"isa\t: " ); +seq_puts (f,"isa\t\t: " );seq_write (f, isa,5 ); isa +=5 ;@@ -96,6 +96,7 @@ static void print_isa(struct seq_file *f, const char *orig_isa) isa++; } } +seq_puts (f,"\n" );/* * If we were given an unsupported ISA in the device tree then print @@ -116,7 +117,7 @@ static void print_mmu(struct seq_file *f, const char *mmu_type)return ;#endif -seq_printf (f,"mmu\t: %s\n" , mmu_type+6 ); +seq_printf (f,"mmu\t\t: %s\n" , mmu_type+6 ); }static void *c_start (struct seq_file *m,loff_t *pos)@@ -144,14 +145,15 @@ static int c_show(struct seq_file *m, void *v) NULL);const char *compat, *isa, *mmu; -seq_printf (m,"hart\t: %lu\n" , cpu_id); +seq_printf (m,"processor\t: %lu\n" , cpu_id); +seq_printf (m,"hart\t\t: %lu\n" ,cpuid_to_hardid_map (cpu_id));if (!of_property_read_string (node,"riscv,isa" , &isa))print_isa (m, isa);if (!of_property_read_string (node,"mmu-type" , &mmu))print_mmu (m, mmu);if (!of_property_read_string (node,"compatible" , &compat) &&strcmp (compat,"riscv" )) -seq_printf (m,"uarch\t: %s\n" , compat); +seq_printf (m,"uarch\t\t: %s\n" , compat);seq_puts (m,"\n" );return 0 ;
Quoted 9 lines · Atish Patra
Quoted 4 lines · Anup Patel
Currently,
/proc/cpuinfoshow logical CPU ID as Hart ID which is in-correct. This patch shows CPU ID and Hart ID separately in/proc/cpuinfousingcpuid_to_hardid_map().I noticed it should be
cpuid_to_hartid_mapinstead ofcpuid_to_hardid_map. It was a typo in my smp cleanup patch series. Sorry for the inconvenience here.
Thanks for pointing.
Quoted 2 lines
I can include this patch in into my series fixing the typo if you want.
Yes, please.
You can also include "RISC-V: Show IPI stats" patch in your series. This patch was reviewed by Christoph.
Thanks,
Anup
Quoted 11 lines · Anup Patel
Quoted 9 lines · Atish Patra
Quoted 4 lines · Anup Patel
Currently,
/proc/cpuinfoshow logical CPU ID as Hart ID which is in-correct. This patch shows CPU ID and Hart ID separately in/proc/cpuinfousingcpuid_to_hardid_map().I noticed it should be
cpuid_to_hartid_mapinstead ofcpuid_to_hardid_map. It was a typo in my smp cleanup patch series. Sorry for the inconvenience here.
Quoted 7 lines
Thanks for pointing.
Quoted 2 lines
I can include this patch in into my series fixing the typo if you want.
Yes, please.
Great.
Quoted 3 lines
You can also include "RISC-V: Show IPI stats" patch in your series. This patch was reviewed by Christoph.
I already had that included as a part of v5.
Regards,
Atish
Quoted 3 lines
Thanks,
Anup