[PATCH v2 0/1] show 'last online CPU' error in dlpar_cpu_offline()

[PATCH v2 0/1] show 'last online CPU' error in dlpar_cpu_offline()

[Daniel Henrique Barboza]

changes in v2 after Michael Ellerman review:
- moved the verification code from dlpar_cpu_remove() to
  dlpar_cpu_offline(), while holding cpu_add_remove_lock
- reworded the commit message and code comment
v1 link: 
https://patchwork.ozlabs.org/project/linuxppc-dev/patch/20210305173845.451158-1-danielhb413@gmail.com/

Daniel Henrique Barboza (1):
  hotplug-cpu.c: show 'last online CPU' error in dlpar_cpu_offline()

 arch/powerpc/platforms/pseries/hotplug-cpu.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

-- 
2.30.2

[PATCH v2 1/1] hotplug-cpu.c: show 'last online CPU' error in dlpar_cpu_offline()

[Daniel Henrique Barboza]

One of the reasons that dlpar_cpu_offline can fail is when attempting to
offline the last online CPU of the kernel. This can be observed in a
pseries QEMU guest that has hotplugged CPUs. If the user offlines all
other CPUs of the guest, and a hotplugged CPU is now the last online
CPU, trying to reclaim it will fail. See [1] for an example.

The current error message in this situation returns rc with -EBUSY and a
generic explanation, e.g.:

pseries-hotplug-cpu: Failed to offline CPU PowerPC,POWER9, rc: -16

EBUSY can be caused by other conditions, such as cpu_hotplug_disable
being true. Throwing a more specific error message for this case,
instead of just "Failed to offline CPU", makes it clearer that the error
is in fact a known error situation instead of other generic/unknown
cause.

This patch adds a 'last online' check in dlpar_cpu_offline() to catch
the 'last online CPU' offline error, eturning a more informative error
message:

pseries-hotplug-cpu: Unable to remove last online CPU PowerPC,POWER9

[1] https://bugzilla.redhat.com/1911414

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
---
 arch/powerpc/platforms/pseries/hotplug-cpu.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/arch/powerpc/platforms/pseries/hotplug-cpu.c b/arch/powerpc/platforms/pseries/hotplug-cpu.c
index 12cbffd3c2e3..3ac7e904385c 100644
--- a/arch/powerpc/platforms/pseries/hotplug-cpu.c
+++ b/arch/powerpc/platforms/pseries/hotplug-cpu.c
@@ -271,6 +271,18 @@ static int dlpar_offline_cpu(struct device_node *dn)
 			if (!cpu_online(cpu))
 				break;
 
+			/* device_offline() will return -EBUSY (via cpu_down())
+			 * if there is only one CPU left. Check it here to fail
+			 * earlier and with a more informative error message,
+			 * while also retaining the cpu_add_remove_lock to be sure
+			 * that no CPUs are being online/offlined during this
+			 * check. */
+			if (num_online_cpus() == 1) {
+				pr_warn("Unable to remove last online CPU %pOFn\n", dn);
+				rc = -EBUSY;
+				goto out_unlock;
+			}
+
 			cpu_maps_update_done();
 			rc = device_offline(get_cpu_device(cpu));
 			if (rc)
@@ -283,6 +295,7 @@ static int dlpar_offline_cpu(struct device_node *dn)
 				thread);
 		}
 	}
+out_unlock:
 	cpu_maps_update_done();
 
 out:
-- 
2.30.2

Re: [PATCH v2 1/1] hotplug-cpu.c: show 'last online CPU' error in dlpar_cpu_offline()

[Daniel Axtens]

Hi Daniel,

Two small nitpicks:

> This patch adds a 'last online' check in dlpar_cpu_offline() to catch
> the 'last online CPU' offline error, eturning a more informative error
                                       ^--- s/eturning/returning/;


> +			/* device_offline() will return -EBUSY (via cpu_down())
> +			 * if there is only one CPU left. Check it here to fail
> +			 * earlier and with a more informative error message,
> +			 * while also retaining the cpu_add_remove_lock to be sure
> +			 * that no CPUs are being online/offlined during this
> +			 * check. */

Checkpatch has a small issue with this comment:

WARNING: Block comments use a trailing */ on a separate line
#50: FILE: arch/powerpc/platforms/pseries/hotplug-cpu.c:279:
+			 * check. */

Apart from that, this patch seems sane to me, but I haven't been able to
test it.

Kind regards,
Daniel

Re: [PATCH v2 1/1] hotplug-cpu.c: show 'last online CPU' error in dlpar_cpu_offline()

[Daniel Henrique Barboza]

Hey Daniel,

On 3/26/21 2:24 AM, Daniel Axtens wrote:
> Hi Daniel,
> 
> Two small nitpicks:
> 
>> This patch adds a 'last online' check in dlpar_cpu_offline() to catch
>> the 'last online CPU' offline error, eturning a more informative error
>                                         ^--- s/eturning/returning/;
> 
> 
>> +			/* device_offline() will return -EBUSY (via cpu_down())
>> +			 * if there is only one CPU left. Check it here to fail
>> +			 * earlier and with a more informative error message,
>> +			 * while also retaining the cpu_add_remove_lock to be sure
>> +			 * that no CPUs are being online/offlined during this
>> +			 * check. */
> 
> Checkpatch has a small issue with this comment:
> 
> WARNING: Block comments use a trailing */ on a separate line
> #50: FILE: arch/powerpc/platforms/pseries/hotplug-cpu.c:279:
> +			 * check. */
> 
> Apart from that, this patch seems sane to me, but I haven't been able to
> test it.


Thanks for the review, and for letting me know of the existence of
'scripts/checkpatch.pl' to verify the patches before posting. I'll
send a v3.


Thanks,


DHB


> 
> Kind regards,
> Daniel
> 

Re: [PATCH v2 0/1] show 'last online CPU' error in dlpar_cpu_offline()

[Michael Ellerman]

On Tue, 23 Mar 2021 17:50:55 -0300, Daniel Henrique Barboza wrote:
> changes in v2 after Michael Ellerman review:
> - moved the verification code from dlpar_cpu_remove() to
>   dlpar_cpu_offline(), while holding cpu_add_remove_lock
> - reworded the commit message and code comment
> v1 link:
> https://patchwork.ozlabs.org/project/linuxppc-dev/patch/20210305173845.451158-1-danielhb413@gmail.com/
> 
> [...]

Applied to powerpc/next.

[1/1] hotplug-cpu.c: show 'last online CPU' error in dlpar_cpu_offline()
      https://git.kernel.org/powerpc/c/d19b3ad02c2d1a9a697b7059e32fa2d97a420b15

cheers