| Message ID | 20200226123844.29519-1-roger.pau@citrix.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | None | expand |
On 26.02.2020 13:38, Roger Pau Monne wrote: > Using scratch_cpumask in send_IPI_mask is not safe in IRQ or exception > context because it can nest, and hence send_IPI_mask could be > overwriting another user scratch cpumask data when used in such > contexts. > > Instead introduce a new cpumask to be used by send_IPI_mask, and > disable interrupts while using it. With this now apparently stale sentence dropped (easily done while committing) > Fallback to not using the scratch cpumask (and hence not attemping to > optimize IPI sending by using a shorthand) when in IRQ or exception > context. Note that the scratch cpumask cannot be used when > non-maskable interrupts are being serviced (NMI or #MC) and hence > fallback to not using the shorthand in that case, like it was done > previously. > > Fixes: 5500d265a2a8 ('x86/smp: use APIC ALLBUT destination shorthand when possible') > Reported-by: Sander Eikelenboom <linux@eikelenboom.it> > Signed-off-by: Roger Pau Monné <roger.pau@citrix.com> Reviewed-by: Jan Beulich <jbeulich@suse.com> Jan
On Wed, Feb 26, 2020 at 02:10:44PM +0100, Jan Beulich wrote: > On 26.02.2020 13:38, Roger Pau Monne wrote: > > Using scratch_cpumask in send_IPI_mask is not safe in IRQ or exception > > context because it can nest, and hence send_IPI_mask could be > > overwriting another user scratch cpumask data when used in such > > contexts. > > > > Instead introduce a new cpumask to be used by send_IPI_mask, and > > disable interrupts while using it. > > With this now apparently stale sentence dropped (easily done > while committing) Uh, I thought I fixed the commit message, but looks like I missed that bit. Thanks.
diff --git a/xen/arch/x86/smp.c b/xen/arch/x86/smp.c index 55d08c9d52..0461812cf6 100644 --- a/xen/arch/x86/smp.c +++ b/xen/arch/x86/smp.c @@ -69,6 +69,18 @@ void send_IPI_mask(const cpumask_t *mask, int vector) bool cpus_locked = false; cpumask_t *scratch = this_cpu(scratch_cpumask); + if ( in_irq() || in_mce_handler() || in_nmi_handler() ) + { + /* + * When in IRQ, NMI or #MC context fallback to the old (and simpler) + * IPI sending routine, and avoid doing any performance optimizations + * (like using a shorthand) in order to avoid using the scratch + * cpumask which cannot be used in interrupt context. + */ + alternative_vcall(genapic.send_IPI_mask, mask, vector); + return; + } + /* * This can only be safely used when no CPU hotplug or unplug operations * are taking place, there are no offline CPUs (unless those have been
Using scratch_cpumask in send_IPI_mask is not safe in IRQ or exception context because it can nest, and hence send_IPI_mask could be overwriting another user scratch cpumask data when used in such contexts. Instead introduce a new cpumask to be used by send_IPI_mask, and disable interrupts while using it. Fallback to not using the scratch cpumask (and hence not attemping to optimize IPI sending by using a shorthand) when in IRQ or exception context. Note that the scratch cpumask cannot be used when non-maskable interrupts are being serviced (NMI or #MC) and hence fallback to not using the shorthand in that case, like it was done previously. Fixes: 5500d265a2a8 ('x86/smp: use APIC ALLBUT destination shorthand when possible') Reported-by: Sander Eikelenboom <linux@eikelenboom.it> Signed-off-by: Roger Pau Monné <roger.pau@citrix.com> --- Changes since v4: - Add _handler suffix to in_nmi/in_mce calls. Changes since v3: - Do not use a dedicated cpumask, and instead prevent usage when in IRQ context. Changes since v2: - Fallback to the previous IPI sending mechanism in #MC or #NMI context. Changes since v1: - Don't use the shorthand when in #MC or #NMI context. --- xen/arch/x86/smp.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)