powerpc/kdump: Fix race in kdump shutdown
Commit: 60adec6226bbcf061d4c2d10944fced209d1847d upstream
When we are crashing, the crashing/primary CPU IPIs the secondaries to
turn off IRQs, go into real mode and wait in kexec_wait. While this
is happening, the primary tears down all the MMU maps. Unfortunately
the primary doesn't check to make sure the secondaries have entered
real mode before doing this.
On PHYP machines, the secondaries can take a long time shutting down
the IRQ controller as RTAS calls are need. These RTAS calls need to
be serialised which resilts in the secondaries contending in
lock_rtas() and hence taking a long time to shut down.
We've hit this on large POWER7 machines, where some secondaries are
still waiting in lock_rtas(), when the primary tears down the HPTEs.
This patch makes sure all secondaries are in real mode before the
primary tears down the MMU. It uses the new kexec_state entry in the
paca. It times out if the secondaries don't reach real mode after
10sec.
Signed-off-by: Michael Neuling <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
Signed-off-by: Kamalesh Babulal <[email protected]>
cc: Anton Blanchard <[email protected]>
---
arch/powerpc/kernel/crash.c | 28 ++++++++++++++++++++++++++++
1 files changed, 28 insertions(+), 0 deletions(-)
diff --git a/arch/powerpc/kernel/crash.c b/arch/powerpc/kernel/crash.c
index b779818..706b622 100644
--- a/arch/powerpc/kernel/crash.c
+++ b/arch/powerpc/kernel/crash.c
@@ -153,6 +153,7 @@ static void crash_kexec_prepare_cpus(int cpu)
while (cpus_weight(cpus_in_crash) < ncpus)
cpu_relax();
}
+
/*
* Make sure all CPUs are entered via soft-reset if the kdump is
* invoked using soft-reset.
@@ -162,6 +163,32 @@ static void crash_kexec_prepare_cpus(int cpu)
/* Leave the IPI callback set */
}
+/* wait for all the CPUs to hit real mode but timeout if they don't come in */
+static void crash_kexec_wait_realmode(int cpu)
+{
+ unsigned int msecs;
+ int i;
+
+ msecs = 10000;
+ for (i = 0; i < NR_CPUS && msecs > 0; i++) {
+ if (i == cpu)
+ continue;
+
+ while (paca[i].kexec_state < KEXEC_STATE_REAL_MODE) {
+ barrier();
+ if (!cpu_possible(i))
+ break;
+
+ if (!cpu_online(i))
+ break;
+
+ msecs--;
+ mdelay(1);
+ }
+ }
+ mb();
+}
+
/*
* This function will be called by secondary cpus or by kexec cpu
* if soft-reset is activated to stop some CPUs.
@@ -419,6 +446,7 @@ void default_machine_crash_shutdown(struct pt_regs *regs)
crash_kexec_prepare_cpus(crashing_cpu);
cpu_set(crashing_cpu, cpus_in_crash);
crash_kexec_stop_spus();
+ crash_kexec_wait_realmode(crashing_cpu);
if (ppc_md.kexec_cpu_down)
ppc_md.kexec_cpu_down(1, 0);
}
On Sat, Mar 19, 2011 at 11:44:33PM +0530, Kamalesh Babulal wrote:
> powerpc/kdump: Fix race in kdump shutdown
>
> Commit: 60adec6226bbcf061d4c2d10944fced209d1847d upstream
>
> When we are crashing, the crashing/primary CPU IPIs the secondaries to
> turn off IRQs, go into real mode and wait in kexec_wait. While this
> is happening, the primary tears down all the MMU maps. Unfortunately
> the primary doesn't check to make sure the secondaries have entered
> real mode before doing this.
>
> On PHYP machines, the secondaries can take a long time shutting down
> the IRQ controller as RTAS calls are need. These RTAS calls need to
> be serialised which resilts in the secondaries contending in
> lock_rtas() and hence taking a long time to shut down.
>
> We've hit this on large POWER7 machines, where some secondaries are
> still waiting in lock_rtas(), when the primary tears down the HPTEs.
>
> This patch makes sure all secondaries are in real mode before the
> primary tears down the MMU. It uses the new kexec_state entry in the
> paca. It times out if the secondaries don't reach real mode after
> 10sec.
>
> Signed-off-by: Michael Neuling <[email protected]>
> Signed-off-by: Benjamin Herrenschmidt <[email protected]>
> Signed-off-by: Kamalesh Babulal <[email protected]>
> cc: Anton Blanchard <[email protected]>
Please don't indent these, I have to unindent them for every patch I
apply from you, which is a pain.
> ---
> arch/powerpc/kernel/crash.c | 28 ++++++++++++++++++++++++++++
> 1 files changed, 28 insertions(+), 0 deletions(-)
>
> diff --git a/arch/powerpc/kernel/crash.c b/arch/powerpc/kernel/crash.c
> index b779818..706b622 100644
> --- a/arch/powerpc/kernel/crash.c
> +++ b/arch/powerpc/kernel/crash.c
> @@ -153,6 +153,7 @@ static void crash_kexec_prepare_cpus(int cpu)
> while (cpus_weight(cpus_in_crash) < ncpus)
> cpu_relax();
> }
> +
> /*
> * Make sure all CPUs are entered via soft-reset if the kdump is
> * invoked using soft-reset.
That chunk is not in the upstream commit. Please be more careful.
greg k-h