[PATCH v2] s390/sclp_vt220: Fix console name to match device
Console name reported in /proc/consoles:
ttyS1-W- (EC p )4:65
does not match the char device name:
crw--w1 root root4, 65 May 17 12:18 /dev/ttysclp0
so debian-installer inside a QEMU s390x instance gets confused and fails
to start with the following error:
steal-ctty: No such file or directory
Signed-off-by: Valentin Vidic
---
v2: also update preferred console for VT220 case
arch/s390/kernel/setup.c | 2 +-
drivers/s390/char/sclp_vt220.c | 4 ++--
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index 72134f9f6ff5..3ec6ca9c26c0 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -165,7 +165,7 @@ static void __init set_preferred_console(void)
else if (CONSOLE_IS_3270)
add_preferred_console("tty3270", 0, NULL);
else if (CONSOLE_IS_VT220)
- add_preferred_console("ttyS", 1, NULL);
+ add_preferred_console("ttysclp", 0, NULL);
else if (CONSOLE_IS_HVC)
add_preferred_console("hvc", 0, NULL);
}
diff --git a/drivers/s390/char/sclp_vt220.c b/drivers/s390/char/sclp_vt220.c
index 047f812d1a1c..71ed1bf15598 100644
--- a/drivers/s390/char/sclp_vt220.c
+++ b/drivers/s390/char/sclp_vt220.c
@@ -35,8 +35,8 @@
#define SCLP_VT220_MINOR 65
#define SCLP_VT220_DRIVER_NAME "sclp_vt220"
#define SCLP_VT220_DEVICE_NAME "ttysclp"
-#define SCLP_VT220_CONSOLE_NAME"ttyS"
-#define SCLP_VT220_CONSOLE_INDEX 1 /* console=ttyS1 */
+#define SCLP_VT220_CONSOLE_NAME"ttysclp"
+#define SCLP_VT220_CONSOLE_INDEX 0 /* console=ttysclp0 */
/* Representation of a single write request */
struct sclp_vt220_request {
--
2.20.1
Re: Bester installation image missing virtio net Modulen
On Tue, Apr 27, 2021 at 03:34:15PM +0200, Christian Borntraeger wrote: > That is OK. I will try to test it then on LPAR and find someone that will test > it on z/VM. Please remove the cc stable as I want to have this tested for > a while. Hopefully it only affects the ttysclp0 console, which is not the > default console for z/VM and LPAR. We would at least need to change the > device driver and commands documentation. Anything else? Ok, I was able to build a recent kernel with the patch. debian-installer starts now and console info is as expected: ~ # uname -a Linux (none) 5.12.0+ #21 SMP Tue Apr 27 17:26:32 CEST 2021 s390x GNU/Linux ~ # dmesg | grep console -i [2.591718] Console: colour dummy device 80x25 [2.600725] printk: console [ttysclp0] enabled ~ # cat /proc/consoles ttysclp0 -W- (EC p )4:65 ~ # ls -l /dev/ttysclp0 crw--w1 root root4, 65 Apr 27 16:30 /dev/ttysclp0 -- Valentin
Re: Bester installation image missing virtio net Modulen
On 27.04.21 14:40, Valentin Vidic wrote: On Tue, Apr 27, 2021 at 12:08:01PM +0200, Christian Borntraeger wrote: I think we did not tell him to go away. It was just that the patch itself has some downsides in that it can break existing howtos (that do contain console=tty0). We could revisit the initial patch (I think we also need to change arch/s390/kernel/setup.c set_preferred_console) and then we would need to test all possible combinations on LPAR, KVM, z/VM and see if we could fixup existing documentation. Would Valentin be willing to provide a new version that also touches setup.c? Sure, I can try to update that patch, but would probably only be able to test it with KVM. That is OK. I will try to test it then on LPAR and find someone that will test it on z/VM. Please remove the cc stable as I want to have this tested for a while. Hopefully it only affects the ttysclp0 console, which is not the default console for z/VM and LPAR. We would at least need to change the device driver and commands documentation. Anything else?
Re: Bester installation image missing virtio net Modulen
On Tue, Apr 27, 2021 at 12:08:01PM +0200, Christian Borntraeger wrote: > I think we did not tell him to go away. It was just that the patch itself > has some downsides in that it can break existing howtos (that do contain > console=tty0). > > We could revisit the initial patch (I think we also need to change > arch/s390/kernel/setup.c set_preferred_console) and then > we would need to test all possible combinations on LPAR, KVM, z/VM > and see if we could fixup existing documentation. > > Would Valentin be willing to provide a new version that also touches > setup.c? Sure, I can try to update that patch, but would probably only be able to test it with KVM. -- Valentin
Re: Bester installation image missing virtio net Modulen
On 27.04.21 09:22, Philipp Kern wrote: On 26.04.21 10:52, Benjamin Zimmermann wrote: $ sudo virt-install [...] --extra-args 'console=ttysclp0 [...] I do see the 'steal_ctty failure' otherwise. I'd be cool if someone from IBM could actually go and fix [1] in some form or another. Valentin tried[2] to do that in the kernel and got told off. Maybe the answer is to bring this to linux-390 but I don't actually know. I think we did not tell him to go away. It was just that the patch itself has some downsides in that it can break existing howtos (that do contain console=tty0). We could revisit the initial patch (I think we also need to change arch/s390/kernel/setup.c set_preferred_console) and then we would need to test all possible combinations on LPAR, KVM, z/VM and see if we could fixup existing documentation. Would Valentin be willing to provide a new version that also touches setup.c? Or, as an alternative we could revisit Valentins initial patch for the installer. I think technically what we want is "just" for /proc/consoles to do the right thing or have a way to map console name to tty name. Kind regards Philipp Kern [1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=961056 [2] https://lkml.org/lkml/2020/5/19/854
Re: Bester installation image missing virtio net Modulen
On 26.04.21 10:52, Benjamin Zimmermann wrote: $ sudo virt-install [...] --extra-args 'console=ttysclp0 [...] > > I do see the 'steal_ctty failure' otherwise. I'd be cool if someone from IBM could actually go and fix [1] in some form or another. Valentin tried[2] to do that in the kernel and got told off. Maybe the answer is to bring this to linux-390 but I don't actually know. I think technically what we want is "just" for /proc/consoles to do the right thing or have a way to map console name to tty name. Kind regards Philipp Kern [1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=961056 [2] https://lkml.org/lkml/2020/5/19/854
