[PATCH 0/1] RFC: Make thunderbolt NHI driver work with kexec [RESEND]

[PATCH 0/1] RFC: Make thunderbolt NHI driver work with kexec [RESEND]

[Maxim Levitsky]

While trying to use kexec on my desktop,
which has Titan Ridge Thunderbolt add-on card,
I noticed that after a kexec, system hangs
for about a minute in the 'wating for udev devices
to settle' phase of the boot.

I found out that if I unload the thunderbolt driver
prior to the kexec, then it works just fine.

Looking at the code I see that the driver doeesn't
implement .shutdown method, so this on-liner patch
points it to the same method that is used on device removal.

I don't know if this is the right solution,
but it does appear to work just fine with this patch.

What do you think?

Best regards,
	Maxim Levitsky

Maxim Levitsky (1):
  thunderbolt: add trivial .shutdown

 drivers/thunderbolt/nhi.c | 1 +
 1 file changed, 1 insertion(+)

-- 
2.25.4

[PATCH 1/1] thunderbolt: add trivial .shutdown

[Maxim Levitsky]

On my machine, a kexec with this driver loaded in the old kernel
causes a very long delay on boot in the kexec'ed kernel,
most likely due to unclean shutdown prior to that.

Unloading thunderbolt driver prior to kexec allows kexec to work as fast
as regular kernel boot, as well as adding this .shutdown pointer.

Shutting a device prior to the shutdown completely is always
a good idea IMHO to help with kexec,
and this one-liner patch implements it.

Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
---
 drivers/thunderbolt/nhi.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/thunderbolt/nhi.c b/drivers/thunderbolt/nhi.c
index 1be491ecbb45..4046642e6aec 100644
--- a/drivers/thunderbolt/nhi.c
+++ b/drivers/thunderbolt/nhi.c
@@ -1285,6 +1285,7 @@ static struct pci_driver nhi_driver = {
 	.id_table = nhi_ids,
 	.probe = nhi_probe,
 	.remove = nhi_remove,
+	.shutdown = nhi_remove,
 	.driver.pm = &nhi_pm_ops,
 };
 
-- 
2.25.4

Re: [PATCH 1/1] thunderbolt: add trivial .shutdown

[Mika Westerberg]

On Wed, May 20, 2020 at 09:52:21PM +0300, Maxim Levitsky wrote:
> On my machine, a kexec with this driver loaded in the old kernel
> causes a very long delay on boot in the kexec'ed kernel,
> most likely due to unclean shutdown prior to that.
> 
> Unloading thunderbolt driver prior to kexec allows kexec to work as fast
> as regular kernel boot, as well as adding this .shutdown pointer.
> 
> Shutting a device prior to the shutdown completely is always
> a good idea IMHO to help with kexec,
> and this one-liner patch implements it.
> 
> Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
> ---
>  drivers/thunderbolt/nhi.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/thunderbolt/nhi.c b/drivers/thunderbolt/nhi.c
> index 1be491ecbb45..4046642e6aec 100644
> --- a/drivers/thunderbolt/nhi.c
> +++ b/drivers/thunderbolt/nhi.c
> @@ -1285,6 +1285,7 @@ static struct pci_driver nhi_driver = {
>  	.id_table = nhi_ids,
>  	.probe = nhi_probe,
>  	.remove = nhi_remove,
> +	.shutdown = nhi_remove,

I guess nhi_remove() does a bit more work than strictly required for
.shutdown but we can deal with that later if needed.

Applied to thunderbolt.git/next, thanks!

[PATCH 0/1] RFC: Make thunderbolt NHI driver work with kexec

[Maxim Levitsky]

While trying to use kexec on my desktop,
which has Titan Ridge Thunderbolt add-on card,
I noticed that after a kexec, system hangs
for about a minute in the 'wating for udev devices
to settle' phase of the boot.

I found out that if I unload the thunderbolt driver
prior to the kexec, then it works just fine.

Looking at the code I see that the driver doeesn't
implement .shutdown method, so this on-liner patch
points it to the same method that is used on device removal.

I don't know if this is the right solution,
but it does appear to work just fine with this patch.

What do you think?

Best regards,
	Maxim Levitsky

Maxim Levitsky (1):
  thunderbolt: add trivial .shutdown

 drivers/thunderbolt/nhi.c | 1 +
 1 file changed, 1 insertion(+)

-- 
2.25.4

[PATCH 1/1] thunderbolt: add trivial .shutdown

[Maxim Levitsky]

On my machine, a kexec with this driver loaded in the old kernel
causes a very long delay on boot in the kexec'ed kernel,
most likely due to unclean shutdown prior to that.

Unloading thunderbolt driver prior to kexec allows kexec to work as fast
as regular kernel boot, as well as adding this .shutdown pointer.

Shutting a device prior to the shutdown completely is always
a good idea IMHO to help with kexec,
and this one-liner patch implements it.

Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
---
 drivers/thunderbolt/nhi.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/thunderbolt/nhi.c b/drivers/thunderbolt/nhi.c
index 1be491ecbb45..4046642e6aec 100644
--- a/drivers/thunderbolt/nhi.c
+++ b/drivers/thunderbolt/nhi.c
@@ -1285,6 +1285,7 @@ static struct pci_driver nhi_driver = {
 	.id_table = nhi_ids,
 	.probe = nhi_probe,
 	.remove = nhi_remove,
+	.shutdown = nhi_remove,
 	.driver.pm = &nhi_pm_ops,
 };
 
-- 
2.25.4

Re: [PATCH 1/1] thunderbolt: add trivial .shutdown

[Maxim Levitsky]

On Wed, 2020-05-20 at 21:12 +0300, Maxim Levitsky wrote:
> On my machine, a kexec with this driver loaded in the old kernel
> causes a very long delay on boot in the kexec'ed kernel,
> most likely due to unclean shutdown prior to that.
> 
> Unloading thunderbolt driver prior to kexec allows kexec to work as fast
> as regular kernel boot, as well as adding this .shutdown pointer.
> 
> Shutting a device prior to the shutdown completely is always
> a good idea IMHO to help with kexec,
> and this one-liner patch implements it.
> 
> Signed-off-by: Maxim Levitsky <mlevitsk@redhat.com>
> ---
>  drivers/thunderbolt/nhi.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/thunderbolt/nhi.c b/drivers/thunderbolt/nhi.c
> index 1be491ecbb45..4046642e6aec 100644
> --- a/drivers/thunderbolt/nhi.c
> +++ b/drivers/thunderbolt/nhi.c
> @@ -1285,6 +1285,7 @@ static struct pci_driver nhi_driver = {
>  	.id_table = nhi_ids,
>  	.probe = nhi_probe,
>  	.remove = nhi_remove,
> +	.shutdown = nhi_remove,
>  	.driver.pm = &nhi_pm_ops,
>  };
>  
Oops, I see that I posted this little fix on a wrong mailing list.
I didn't update the script correctly.
Sorry for the noise!

Best regards,
	Maxim Levitsky