hw/s390x: Emit a warning if user tried to enable USB
PATCH 1/1 hw/s390x: Emit a warning if user tried to enable USB
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.
hw/s390x/s390-virtio-ccw.c | 4 ++++ 1 file changed, 4 insertions(+)
hw/s390x/s390-virtio-ccw.c +4 -0
@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine)VirtualCssBus *css_bus;DeviceState *dev; +if (machine->usb) { +warn_report ("This machine does not support USB" ); + } +s390_sclp_init ();/* init memory + setup max page size. Required for the CPU model */ s390_memory_init (machine->ram_size);
Signature · 1 line
2.18.1
Quoted 24 lines · Cornelia Huck
Quoted 21 lines · On Thu, 17 Oct 2019 16:21:23 +0200 Thomas Huth <[email protected]> wrote:
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.hw/s390x/s390-virtio-ccw.c | 4 ++++ 1 file changed, 4 insertions(+)hw/s390x/s390-virtio-ccw.c +0
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine)VirtualCssBus *css_bus;DeviceState *dev;
Quoted 5 lines
hw/s390x/s390-virtio-ccw.c static void ccw_init(MachineState *machine) · line 247
Quoted 2 lines
hw/s390x/s390-virtio-ccw.c +2
-0+if (machine->usb) { +warn_report ("This machine does not support USB" );I'm wondering if this is the only machine type not supporting usb... if not, how are others handling it?
I think most machines are silently ignoring it, like we did on s390x until now, too.
Quoted 2 lines
The usb parsing code in
machine.cdoes not care if usb is even configured (CONFIG_USB).
machine.c is common code, so you can not use CONFIG_USB there.
Quoted 3 lines
There's other stuff in there like
igd-passthru, which seems to be x86 specific; probably historical reasons?
IMHO igd-passthru should be moved to the xen machine, which seems to be the only user.
Thomas
Quoted 18 lines · On Thu, 17 Oct 2019 16:21:23 +0200 Thomas Huth <[email protected]> wrote:
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.hw/s390x/s390-virtio-ccw.c | 4 ++++ 1 file changed, 4 insertions(+)hw/s390x/s390-virtio-ccw.c +0
-0
hw/s390x/s390-virtio-ccw.c static void ccw_init(MachineState *machine) · line 247
Quoted 5 lines
hw/s390x/s390-virtio-ccw.c +2
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine)VirtualCssBus *css_bus;DeviceState *dev; +if (machine->usb) { +warn_report ("This machine does not support USB" );
I'm wondering if this is the only machine type not supporting usb... if not, how are others handling it?
The usb parsing code in machine.c does not care if usb is even configured (CONFIG_USB). There's other stuff in there like igd-passthru, which seems to be x86 specific; probably historical reasons?
Quoted 5 lines
hw/s390x/s390-virtio-ccw.c +2
-0+ } +s390_sclp_init ();/* init memory + setup max page size. Required for the CPU model */ s390_memory_init (machine->ram_size);
Quoted 9 lines · Thomas Huth
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.
I think a warn message is ok, but we should never make this a hard error.
I am pretty sure that there are some tools in the wild that create xmls or qemu commands lines cross-platform and deploy those dynamically. These tools have probably been fixed to work good enough with s390x but nobody with qemu clue has ever looked at these command lines. And I am pretty sure that no user will actually see the command like nor the error message.
So this warning will stay unnoticed until we make this a hard error. And then we have broken a previously working setup.
In other words, I appreciate the willingness to detect mis-uses but I fear that we will never be able to assume that everything is fixed.
Quoted 20 lines
hw/s390x/s390-virtio-ccw.c | 4 ++++ 1 file changed, 4 insertions(+)hw/s390x/s390-virtio-ccw.c +4
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine)VirtualCssBus *css_bus;DeviceState *dev; +if (machine->usb) { +warn_report ("This machine does not support USB" ); + } +s390_sclp_init ();/* init memory + setup max page size. Required for the CPU model */ s390_memory_init (machine->ram_size);
Quoted 25 lines · On Thu, 17 Oct 2019 16:40:56 +0200 Thomas Huth <[email protected]> wrote:
Quoted 24 lines · Cornelia Huck
Quoted 21 lines · On Thu, 17 Oct 2019 16:21:23 +0200 Thomas Huth <[email protected]> wrote:
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.hw/s390x/s390-virtio-ccw.c | 4 ++++ 1 file changed, 4 insertions(+)hw/s390x/s390-virtio-ccw.c +0
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine)VirtualCssBus *css_bus;DeviceState *dev;
Quoted 8 lines
Quoted 5 lines
hw/s390x/s390-virtio-ccw.c static void ccw_init(MachineState *machine) · line 247
Quoted 2 lines
hw/s390x/s390-virtio-ccw.c +2
-0+if (machine->usb) { +warn_report ("This machine does not support USB" );I'm wondering if this is the only machine type not supporting usb... if not, how are others handling it?
I think most machines are silently ignoring it, like we did on s390x until now, too.
I'm wondering how many options we have that do nothing when configured :)
Quoted 5 lines
Quoted 3 lines
The usb parsing code in
machine.cdoes not care if usb is even configured (CONFIG_USB).
machine.cis common code, so you can not useCONFIG_USBthere.
Hm, yes.
Quoted 7 lines
Quoted 4 lines
There's other stuff in there like
igd-passthru, which seems to be x86 specific; probably historical reasons?IMHO igd-passthru should be moved to the xen machine, which seems to be the only user.
OTOH, I can currently specify igd-passthru=on on any machine, which would break if we moved it. Not that it makes any sense...
Quoted 23 lines · Christian Borntraeger
Quoted 9 lines · Thomas Huth
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.I think a warn message is ok, but we should never make this a hard error.
I am pretty sure that there are some tools in the wild that create xmls or qemu commands lines cross-platform and deploy those dynamically. These tools have probably been fixed to work good enough with s390x but nobody with qemu clue has ever looked at these command lines. And I am pretty sure that no user will actually see the command like nor the error message.
Quoted 5 lines
So this warning will stay unnoticed until we make this a hard error. And then we have broken a previously working setup.
In other words, I appreciate the willingness to detect mis-uses but I fear that we will never be able to assume that everything is fixed.
That's what the deprecation process is for. This patch is incomplete unless it also touches qemu-deprecated.texi.
Signature · 3 lines
Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3226 Virtualization: qemu.org | libvirt.org
Quoted 31 lines · Thomas Huth
Quoted 24 lines · Cornelia Huck
Quoted 21 lines · On Thu, 17 Oct 2019 16:21:23 +0200 Thomas Huth <[email protected]> wrote:
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.hw/s390x/s390-virtio-ccw.c | 4 ++++ 1 file changed, 4 insertions(+)hw/s390x/s390-virtio-ccw.c +0
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine)VirtualCssBus *css_bus;DeviceState *dev;Quoted 5 lines
hw/s390x/s390-virtio-ccw.c static void ccw_init(MachineState *machine) · line 247
Quoted 2 lines
hw/s390x/s390-virtio-ccw.c +2
-0+if (machine->usb) { +warn_report ("This machine does not support USB" );I'm wondering if this is the only machine type not supporting usb... if not, how are others handling it?
Quoted 7 lines
I think most machines are silently ignoring it, like we did on s390x until now, too.
Quoted 2 lines
The usb parsing code in
machine.cdoes not care if usb is even configured (CONFIG_USB).
machine.cis common code, so you can not useCONFIG_USBthere.
We already have:
bool target_words_bigendian(void)
{
#if defined(TARGET_WORDS_BIGENDIAN)
return true;
#else
return false;#endif
}
What about something such:
-- >8 --
hw/core/machine.c +1 -1
@@ -986,7 +986,7 @@ static void machine_finalize(Object *obj)bool machine_usb (MachineState *machine) { -return machine->usb; +return machine_has_usb () && machine->usb; }bool machine_kernel_irqchip_allowed (MachineState *machine)
hw/usb/Makefile.objs +1 -0
@@ -1,6 +1,7 @@ # usb subsystem core common-obj-y += core.o combined-packet.o bus.o libhw.o common-obj-$(CONFIG_USB) += desc.o desc-msos.o +obj-y += machine.o # usb host adapters common-obj-$(CONFIG_USB_UHCI) += hcd-uhci.o
hw/usb/machine.c new file +12 -0
@@ -0,0 +1,12 @@ +#include "qemu/osdep.h" +#include "hw/boards.h" +#include "config-devices.h" + +bool machine_has_usb (void ) +{ +#if defined(CONFIG_USB ) +return true; +#else +return false; +#endif +}
include/hw/boards.h +1 -0
@@ -63,6 +63,7 @@ extern MachineState *current_machine;void machine_run_board_init (MachineState *machine);bool machine_usb (MachineState *machine); +bool machine_has_usb (void );/* or target_has_usb()? */ bool machine_kernel_irqchip_allowed (MachineState *machine);bool machine_kernel_irqchip_required (MachineState *machine);bool machine_kernel_irqchip_split (MachineState *machine);
---
Quoted 34 lines · Cornelia Huck
Quoted 31 lines · On Thu, 17 Oct 2019 16:40:56 +0200 Thomas Huth <[email protected]> wrote:
Quoted 29 lines · Cornelia Huck
Quoted 18 lines · On Thu, 17 Oct 2019 16:21:23 +0200 Thomas Huth <[email protected]> wrote:
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.hw/s390x/s390-virtio-ccw.c | 4 ++++ 1 file changed, 4 insertions(+)hw/s390x/s390-virtio-ccw.c +0
-0hw/s390x/s390-virtio-ccw.c static void ccw_init(MachineState *machine) · line 247
Quoted 5 lines
hw/s390x/s390-virtio-ccw.c +2
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine)VirtualCssBus *css_bus;DeviceState *dev; +if (machine->usb) { +warn_report ("This machine does not support USB" );I'm wondering if this is the only machine type not supporting usb... if not, how are others handling it?
Quoted 5 lines
Quoted 2 lines
I think most machines are silently ignoring it, like we did on s390x until now, too.
I'm wondering how many options we have that do nothing when configured :)
Plenty.
-machine usb=on (and its sugared form -usb) are just one of many options that deposit configuration for the board to pick up. Some boards do, some don't. Some boards pick up and reject some configuration they can't use. Some don't. It's a big family of hack jobs.
For -drive (and its sugared forms), we have generic code to turn "not picked up" into an error: drive_check_orphaned(). a66c9dc734 "blockdev: Orphaned drive search", 2014-10-03 720b8dc052 "blockdev: Make orphaned -drive fatal", 2017-02-21
[...]
Quoted 41 lines · Philippe Mathieu-Daudé
Quoted 30 lines · Thomas Huth
Quoted 16 lines · Cornelia Huck
Quoted 13 lines · On Thu, 17 Oct 2019 16:21:23 +0200 Thomas Huth <[email protected]> wrote:
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.
hw/s390x/s390-virtio-ccw.c| 4 ++++1 file changed, 4 insertions(+)Quoted 12 lines
hw/s390x/s390-virtio-ccw.c static void ccw_init(MachineState *machine) · line 246
Quoted 9 lines
hw/s390x/s390-virtio-ccw.c +1
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine) VirtualCssBus *css_bus; DeviceState *dev; + if (machine->usb) { + warn_report("This machine does not support USB");I'm wondering if this is the only machine type not supporting usb... if not, how are others handling it?
Quoted 7 lines
I think most machines are silently ignoring it, like we did on s390x until now, too.
Quoted 2 lines
The usb parsing code in
machine.cdoes not care if usb is even configured (CONFIG_USB).
machine.cis common code, so you can not useCONFIG_USBthere.We already have:
Quoted 8 lines
bool
target_words_bigendian(void)
{
#ifdefined(TARGET_WORDS_BIGENDIAN)
return true;
#else
return false;
#endif
}
... and kvm_available() and xen_available() ...
Quoted 8 lines
hw/usb/machine.c new file +2
-0@@ -0,0 +1,12 @@ +#include "qemu/osdep.h" +#include "hw/boards.h"
hw/usb/machine.c line 12
Quoted 10 lines
hw/usb/machine.c new file +10
-0+#include "config-devices.h" + +bool machine_has_usb (void ) +{ +#if defined(CONFIG_USB ) + return true; +#else + return false; +#endif +}
I think I'd rather call it usb_available() (like the other _available() functions) and put it into arch_init.c (and rename that file to arch.c or target.c or something like that).
Thomas
Quoted 67 lines · Thomas Huth
Quoted 41 lines · Philippe Mathieu-Daudé
Quoted 30 lines · Thomas Huth
Quoted 16 lines · Cornelia Huck
Quoted 13 lines · On Thu, 17 Oct 2019 16:21:23 +0200 Thomas Huth <[email protected]> wrote:
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.hw/s390x/s390-virtio-ccw.c | 4 ++++ 1 file changed, 4 insertions(+)Quoted 12 lines
hw/s390x/s390-virtio-ccw.c static void ccw_init(MachineState *machine) · line 246
Quoted 9 lines
hw/s390x/s390-virtio-ccw.c +1
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine) VirtualCssBus *css_bus; DeviceState *dev; + if (machine->usb) { + warn_report("This machine does not support USB");I'm wondering if this is the only machine type not supporting usb... if not, how are others handling it?
Quoted 7 lines
I think most machines are silently ignoring it, like we did on s390x until now, too.
Quoted 2 lines
The usb parsing code in
machine.cdoes not care if usb is even configured (CONFIG_USB).
machine.cis common code, so you can not useCONFIG_USBthere.We already have:
Quoted 8 lines
bool
target_words_bigendian(void)
{
#ifdefined(TARGET_WORDS_BIGENDIAN)
return true;
#else
return false;
#endif
}... and
kvm_available()andxen_available()...Quoted 14 lines
hw/usb/machine.c new file +8
-0@@ -0,0 +1,12 @@ +#include "qemu/osdep.h" +#include "hw/boards.h" +#include "config-devices.h" + +bool machine_has_usb (void ) +{ +#if defined(CONFIG_USB ) + return true;
Quoted 8 lines
hw/usb/machine.c line 12
Quoted 4 lines
hw/usb/machine.c new file +4
-0+#else + return false; +#endif +}I think I'd rather call it
usb_available()(like the other_available()functions) and put it intoarch_init.c(and rename that file toarch.cortarget.cor something like that).
Yes, clever names :)
Quoted 67 lines · On Fri, 18 Oct 2019 08:35:17 +0200 Thomas Huth <[email protected]> wrote:
Quoted 41 lines · Philippe Mathieu-Daudé
Quoted 30 lines · Thomas Huth
Quoted 16 lines · Cornelia Huck
Quoted 13 lines · On Thu, 17 Oct 2019 16:21:23 +0200 Thomas Huth <[email protected]> wrote:
There is no USB on s390x, so running qemu-system-s390x with "-machine ...,usb=on" is certainly wrong. Emit a warning to make the users aware of their misconfiguration.
Signed-off-by Thomas Huth <[email protected]>---
After a year or two, we could finally turn this into a hard error, but I think we should give the users some time to fix their command lines first, so I'm initially only emitting a warning here.
hw/s390x/s390-virtio-ccw.c| 4 ++++1 file changed, 4 insertions(+)Quoted 12 lines
hw/s390x/s390-virtio-ccw.c static void ccw_init(MachineState *machine) · line 246
Quoted 9 lines
hw/s390x/s390-virtio-ccw.c +1
-0@@ -243,6 +243,10 @@ static void ccw_init(MachineState *machine) VirtualCssBus *css_bus; DeviceState *dev; + if (machine->usb) { + warn_report("This machine does not support USB");I'm wondering if this is the only machine type not supporting usb... if not, how are others handling it?
Quoted 7 lines
I think most machines are silently ignoring it, like we did on s390x until now, too.
Quoted 2 lines
The usb parsing code in
machine.cdoes not care if usb is even configured (CONFIG_USB).
machine.cis common code, so you can not useCONFIG_USBthere.We already have:
Quoted 8 lines
bool
target_words_bigendian(void)
{
#ifdefined(TARGET_WORDS_BIGENDIAN)
return true;
#else
return false;
#endif
}... and
kvm_available()andxen_available()...Quoted 14 lines
hw/usb/machine.c new file +8
-0@@ -0,0 +1,12 @@ +#include "qemu/osdep.h" +#include "hw/boards.h" +#include "config-devices.h" + +bool machine_has_usb (void ) +{ +#if defined(CONFIG_USB ) + return true;
Quoted 8 lines
hw/usb/machine.c line 12
Quoted 4 lines
hw/usb/machine.c new file +4
-0+#else + return false; +#endif +}I think I'd rather call it
usb_available()(like the other_available()functions) and put it intoarch_init.c(and rename that file toarch.cortarget.cor something like that).
I like 'usb_available()'.
Maybe we should also warn for igd_passthru if not xen_available()? Not sure how helpful that is, though. Even if we warn and put it in the deprecation notes, I'm not sure how much we'd gain if we were to make it an actual error.