| Message ID | 146399210850.26372.5857516772891352049.stgit@bahia.huguette.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
On 23.05.2016 10:28, Greg Kurz wrote: > A strange behaviour is observed when comparing PCI hotplug in QEMU, between > x86 and pseries. If you consider the following steps: > - start a VM > - add a PCI device via the QEMU monitor before the rtasd has started (for > example starting the VM in paused state, or hotplug during FW or boot > loader) > - resume the VM execution > > The x86 kernel detects the PCI device, but the pseries one does not. > > This happens because the rtasd kernel worker is currently started under > device_initcall, while PCI probing happens earlier under subsys_initcall. > > As a consequence, if we have a pending RTAS event at boot time, a message > is printed and the event is dropped. > > This patch moves all the initialization of rtasd to arch_initcall, which is > run before subsys_call: this way, logging_enabled is true when the RTAS > event pops up and it is not lost anymore. > > The proc fs bits stay at device_initcall because they cannot be run before > fs_initcall. > > Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com> > --- > arch/powerpc/kernel/rtasd.c | 19 ++++++++++++++----- > 1 file changed, 14 insertions(+), 5 deletions(-) By the way, same is true for device UNplugging: When unplugging devices in QEMU while the firmware is still running, they are never properly removed from the guest. I've checked it, and your patch fixes this problem as well! Great :-) Tested-by: Thomas Huth <thuth@redhat.com>
On Mon, 23 May 2016 20:23:19 +0200 Thomas Huth <thuth@redhat.com> wrote: > On 23.05.2016 10:28, Greg Kurz wrote: > > A strange behaviour is observed when comparing PCI hotplug in QEMU, between > > x86 and pseries. If you consider the following steps: > > - start a VM > > - add a PCI device via the QEMU monitor before the rtasd has started (for > > example starting the VM in paused state, or hotplug during FW or boot > > loader) > > - resume the VM execution > > > > The x86 kernel detects the PCI device, but the pseries one does not. > > > > This happens because the rtasd kernel worker is currently started under > > device_initcall, while PCI probing happens earlier under subsys_initcall. > > > > As a consequence, if we have a pending RTAS event at boot time, a message > > is printed and the event is dropped. > > > > This patch moves all the initialization of rtasd to arch_initcall, which is > > run before subsys_call: this way, logging_enabled is true when the RTAS > > event pops up and it is not lost anymore. > > > > The proc fs bits stay at device_initcall because they cannot be run before > > fs_initcall. > > > > Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com> > > --- > > arch/powerpc/kernel/rtasd.c | 19 ++++++++++++++----- > > 1 file changed, 14 insertions(+), 5 deletions(-) > > By the way, same is true for device UNplugging: When unplugging devices > in QEMU while the firmware is still running, they are never properly > removed from the guest. I've checked it, and your patch fixes this > problem as well! Great :-) > > Tested-by: Thomas Huth <thuth@redhat.com> > Indeed, all pending RTAS events were lost... now rtasd will log them, up to 64 events, but I did not test that far :) Thanks for testing ! Cheers. -- Greg
On Mon, 23 May 2016 10:28:28 +0200 Greg Kurz <gkurz@linux.vnet.ibm.com> wrote: > A strange behaviour is observed when comparing PCI hotplug in QEMU, between > x86 and pseries. If you consider the following steps: > - start a VM > - add a PCI device via the QEMU monitor before the rtasd has started (for > example starting the VM in paused state, or hotplug during FW or boot > loader) > - resume the VM execution > > The x86 kernel detects the PCI device, but the pseries one does not. > > This happens because the rtasd kernel worker is currently started under > device_initcall, while PCI probing happens earlier under subsys_initcall. > > As a consequence, if we have a pending RTAS event at boot time, a message > is printed and the event is dropped. > > This patch moves all the initialization of rtasd to arch_initcall, which is > run before subsys_call: this way, logging_enabled is true when the RTAS > event pops up and it is not lost anymore. > > The proc fs bits stay at device_initcall because they cannot be run before > fs_initcall. > > Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com> > --- Ping ? > arch/powerpc/kernel/rtasd.c | 19 ++++++++++++++----- > 1 file changed, 14 insertions(+), 5 deletions(-) > > diff --git a/arch/powerpc/kernel/rtasd.c b/arch/powerpc/kernel/rtasd.c > index e864b7c5884e..ad9e4e1a2d5d 100644 > --- a/arch/powerpc/kernel/rtasd.c > +++ b/arch/powerpc/kernel/rtasd.c > @@ -526,10 +526,8 @@ void rtas_cancel_event_scan(void) > } > EXPORT_SYMBOL_GPL(rtas_cancel_event_scan); > > -static int __init rtas_init(void) > +static int __init rtas_event_scan_init(void) > { > - struct proc_dir_entry *entry; > - > if (!machine_is(pseries) && !machine_is(chrp)) > return 0; > > @@ -562,13 +560,24 @@ static int __init rtas_init(void) > return -ENOMEM; > } > > + start_event_scan(); > + > + return 0; > +} > +arch_initcall(rtas_event_scan_init); > + > +static int __init rtas_init(void) > +{ > + struct proc_dir_entry *entry; > + > + if (!machine_is(pseries) && !machine_is(chrp)) > + return 0; > + > entry = proc_create("powerpc/rtas/error_log", S_IRUSR, NULL, > &proc_rtas_log_operations); > if (!entry) > printk(KERN_ERR "Failed to create error_log proc entry\n"); > > - start_event_scan(); > - > return 0; > } > __initcall(rtas_init); > >
On Mon, 2016-23-05 at 08:28:28 UTC, Greg Kurz wrote: > A strange behaviour is observed when comparing PCI hotplug in QEMU, between > x86 and pseries. If you consider the following steps: > - start a VM > - add a PCI device via the QEMU monitor before the rtasd has started (for > example starting the VM in paused state, or hotplug during FW or boot > loader) > - resume the VM execution > > The x86 kernel detects the PCI device, but the pseries one does not. > > This happens because the rtasd kernel worker is currently started under > device_initcall, while PCI probing happens earlier under subsys_initcall. > > As a consequence, if we have a pending RTAS event at boot time, a message > is printed and the event is dropped. > > This patch moves all the initialization of rtasd to arch_initcall, which is > run before subsys_call: this way, logging_enabled is true when the RTAS > event pops up and it is not lost anymore. > > The proc fs bits stay at device_initcall because they cannot be run before > fs_initcall. > > Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com> > Tested-by: Thomas Huth <thuth@redhat.com> Has this been tested on PowerVM ? cheers
On Fri, 10 Jun 2016 15:18:32 +1000 (AEST) Michael Ellerman <mpe@ellerman.id.au> wrote: > On Mon, 2016-23-05 at 08:28:28 UTC, Greg Kurz wrote: > > A strange behaviour is observed when comparing PCI hotplug in QEMU, between > > x86 and pseries. If you consider the following steps: > > - start a VM > > - add a PCI device via the QEMU monitor before the rtasd has started (for > > example starting the VM in paused state, or hotplug during FW or boot > > loader) > > - resume the VM execution > > > > The x86 kernel detects the PCI device, but the pseries one does not. > > > > This happens because the rtasd kernel worker is currently started under > > device_initcall, while PCI probing happens earlier under subsys_initcall. > > > > As a consequence, if we have a pending RTAS event at boot time, a message > > is printed and the event is dropped. > > > > This patch moves all the initialization of rtasd to arch_initcall, which is > > run before subsys_call: this way, logging_enabled is true when the RTAS > > event pops up and it is not lost anymore. > > > > The proc fs bits stay at device_initcall because they cannot be run before > > fs_initcall. > > > > Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com> > > Tested-by: Thomas Huth <thuth@redhat.com> > > Has this been tested on PowerVM ? > > cheers > No but I shall do it. Thanks for pointing this out. -- Greg
diff --git a/arch/powerpc/kernel/rtasd.c b/arch/powerpc/kernel/rtasd.c index e864b7c5884e..ad9e4e1a2d5d 100644 --- a/arch/powerpc/kernel/rtasd.c +++ b/arch/powerpc/kernel/rtasd.c @@ -526,10 +526,8 @@ void rtas_cancel_event_scan(void) } EXPORT_SYMBOL_GPL(rtas_cancel_event_scan); -static int __init rtas_init(void) +static int __init rtas_event_scan_init(void) { - struct proc_dir_entry *entry; - if (!machine_is(pseries) && !machine_is(chrp)) return 0; @@ -562,13 +560,24 @@ static int __init rtas_init(void) return -ENOMEM; } + start_event_scan(); + + return 0; +} +arch_initcall(rtas_event_scan_init); + +static int __init rtas_init(void) +{ + struct proc_dir_entry *entry; + + if (!machine_is(pseries) && !machine_is(chrp)) + return 0; + entry = proc_create("powerpc/rtas/error_log", S_IRUSR, NULL, &proc_rtas_log_operations); if (!entry) printk(KERN_ERR "Failed to create error_log proc entry\n"); - start_event_scan(); - return 0; } __initcall(rtas_init);
A strange behaviour is observed when comparing PCI hotplug in QEMU, between x86 and pseries. If you consider the following steps: - start a VM - add a PCI device via the QEMU monitor before the rtasd has started (for example starting the VM in paused state, or hotplug during FW or boot loader) - resume the VM execution The x86 kernel detects the PCI device, but the pseries one does not. This happens because the rtasd kernel worker is currently started under device_initcall, while PCI probing happens earlier under subsys_initcall. As a consequence, if we have a pending RTAS event at boot time, a message is printed and the event is dropped. This patch moves all the initialization of rtasd to arch_initcall, which is run before subsys_call: this way, logging_enabled is true when the RTAS event pops up and it is not lost anymore. The proc fs bits stay at device_initcall because they cannot be run before fs_initcall. Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com> --- arch/powerpc/kernel/rtasd.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-)