Message ID | 1550089092-28783-6-git-send-email-atish.patra@wdc.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | Various SMP related fixes | expand |
On Wed, 13 Feb 2019, Atish Patra wrote: > --- a/arch/riscv/kernel/smpboot.c > +++ b/arch/riscv/kernel/smpboot.c > @@ -66,6 +66,11 @@ void __init setup_smp(void) > found_boot_cpu = 1; > continue; > } > + if (cpuid >= NR_CPUS) { Use nr_cpu_ids instead? Its initialized to NR_CPUS but can be restricted if we can determine on boot how many processor we truly have.
On 2/14/19 9:59 AM, Christopher Lameter wrote: > On Wed, 13 Feb 2019, Atish Patra wrote: > >> --- a/arch/riscv/kernel/smpboot.c >> +++ b/arch/riscv/kernel/smpboot.c >> @@ -66,6 +66,11 @@ void __init setup_smp(void) >> found_boot_cpu = 1; >> continue; >> } >> + if (cpuid >= NR_CPUS) { > > Use nr_cpu_ids instead? Its initialized to NR_CPUS but can be restricted > if we can determine on boot how many processor we truly have. > nr_cpu_ids can be modified by nr_cpus in boot command line. If nr_cpus is set to a value less than number of cpus described in the DT, kernel may not boot depending on the boot hartid. This happens because we break out of the while loop and BUG_ON(!found_boot_cpu). As the required fix will be different from the current fix, I would like to do this in a follow up patch if that's ok. Btw, thanks for bringing this up. Regards, Atish
diff --git a/arch/riscv/kernel/smpboot.c b/arch/riscv/kernel/smpboot.c index d369b669..eb533b5c 100644 --- a/arch/riscv/kernel/smpboot.c +++ b/arch/riscv/kernel/smpboot.c @@ -66,6 +66,11 @@ void __init setup_smp(void) found_boot_cpu = 1; continue; } + if (cpuid >= NR_CPUS) { + pr_warn("Invalid cpuid [%d] for hartid [%d]\n", + cpuid, hart); + break; + } cpuid_to_hartid_map(cpuid) = hart; set_cpu_possible(cpuid, true);