Unbreak make check SPEED=slow
cover Unbreak make check SPEED=slow
On the one hand, getting bitten by this bug in production seems somewhat unlikely. On the other hand, shipping with broken "make check" would be embarrassing. So let's consider the fix for 5.1.
Markus Armbruster (2):
msf2: Unbreak device-list-properties for "msf-soc"
hw: Mark nd_table[] misuse in realize methods FIXME
hw/arm/allwinner-h3.c | 1 + hw/arm/msf2-soc.c | 9 +++++---- hw/arm/xlnx-versal.c | 1 + hw/arm/xlnx-zynqmp.c | 1 + hw/dma/sparc32_dma.c | 1 + hw/riscv/sifive_u.c | 1 + 6 files changed, 10 insertions(+), 4 deletions(-)
Signature · 1 line
2.26.2
PATCH 1/2 msf2: Unbreak device-list-properties for "msf-soc"
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio
{"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}}
{"execute": "qmp_capabilities"}
{"return": {}}
{"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}}
Unsupported NIC model: ftgmac100
armbru@dusky:~/work/images$ echo $?
1This is what breaks "make check SPEED=slow".
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via qemu_check_nic_model(). That's wrong.
We fixed the exact same bug for device "allwinner-a10" in commit 8aabc5437b "hw/arm/allwinner-a10: Do not use nd_table in instance_init function". Fix this instance the same way: move the offending code to m2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.
--- hw/arm/msf2-soc.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
hw/arm/msf2-soc.c +5 -4
@@ -82,10 +82,6 @@ static void m2sxxx_soc_initfn(Object *obj) }object_initialize_child (obj,"emac" , &s->emac,TYPE_MSS_EMAC ); -if (nd_table[0 ].used) { -qemu_check_nic_model (&nd_table[0 ],TYPE_MSS_EMAC ); -qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]); - } }static void m2sxxx_soc_realize (DeviceState *dev_soc,Error **errp)
@@ -187,6 +183,11 @@ static void m2sxxx_soc_realize(DeviceState *dev_soc, Error **errp)g_free (bus_name); } +/* FIXME use qdev NIC properties instead of nd_table[] */ +if (nd_table[0 ].used) { +qemu_check_nic_model (&nd_table[0 ],TYPE_MSS_EMAC ); +qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]); + } dev =DEVICE (&s->emac);object_property_set_link (OBJECT (&s->emac),"ahb-bus" ,OBJECT (get_system_memory ()), &error_abort);
Signature · 1 line
2.26.2
Re: [PATCH for-5.1 1/2] msf2: Unbreak device-list-properties for "msf-soc"
Quoted 13 lines · Markus Armbruster
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio {"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}} {"execute": "qmp_capabilities"} {"return": {}} {"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}} Unsupported NIC model: ftgmac100 armbru@dusky:~/work/images$ echo $? 1This is what breaks "make check SPEED=slow".
Quoted 7 lines
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via
qemu_check_nic_model(). That's wrong.We fixed the exact same bug for device "allwinner-a10" in commit
8aabc5437b"hw/arm/allwinner-a10: Do not usend_tableininstance_initfunction". Fix this instance the same way: move the offending code tom2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.
That addresses this other thread, right?
https://www.mail-archive.com/[email protected]/msg720658.html
Quoted 35 lines
Fixes 05b7374a58cd18aa3516e33513808896d0ac9b7bSigned-off-by Markus Armbruster <[email protected]>--- hw/arm/msf2-soc.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)hw/arm/msf2-soc.c +5
-4@@ -82,10 +82,6 @@ static void m2sxxx_soc_initfn(Object *obj) }object_initialize_child (obj,"emac" , &s->emac,TYPE_MSS_EMAC ); -if (nd_table[0 ].used) { -qemu_check_nic_model (&nd_table[0 ],TYPE_MSS_EMAC ); -qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]); - } }static void m2sxxx_soc_realize (DeviceState *dev_soc,Error **errp)@@ -187,6 +183,11 @@ static void m2sxxx_soc_realize(DeviceState *dev_soc, Error **errp)g_free (bus_name); } +/* FIXME use qdev NIC properties instead of nd_table[] */ +if (nd_table[0 ].used) { +qemu_check_nic_model (&nd_table[0 ],TYPE_MSS_EMAC ); +qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]); + } dev =DEVICE (&s->emac);object_property_set_link (OBJECT (&s->emac),"ahb-bus" ,OBJECT (get_system_memory ()), &error_abort);
Re: [PATCH for-5.1 1/2] msf2: Unbreak device-list-properties for "msf-soc"
Quoted 17 lines · Philippe Mathieu-Daudé
Quoted 16 lines · Markus Armbruster
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio {"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}} {"execute": "qmp_capabilities"} {"return": {}} {"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}} Unsupported NIC model: ftgmac100 armbru@dusky:~/work/images$ echo $? 1This is what breaks "make check SPEED=slow".
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via
qemu_check_nic_model(). That's wrong.
Quoted 7 lines
Quoted 4 lines
We fixed the exact same bug for device "allwinner-a10" in commit
8aabc5437b"hw/arm/allwinner-a10: Do not usend_tableininstance_initfunction". Fix this instance the same way: move the offending code tom2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.That addresses this other thread, right?
https://www.mail-archive.com/[email protected]/msg720658.html
Correct! I wasn't aware of it, thanks for making the connection.
Re: [PATCH for-5.1 1/2] msf2: Unbreak device-list-properties for "msf-soc"
Quoted 25 lines · Markus Armbruster
Quoted 22 lines · Philippe Mathieu-Daudé
Quoted 20 lines · Markus Armbruster
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio {"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}} {"execute": "qmp_capabilities"} {"return": {}} {"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}} Unsupported NIC model: ftgmac100 armbru@dusky:~/work/images$ echo $? 1This is what breaks "make check SPEED=slow".
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via
qemu_check_nic_model(). That's wrong.We fixed the exact same bug for device "allwinner-a10" in commit
8aabc5437b"hw/arm/allwinner-a10: Do not usend_tableininstance_initfunction". Fix this instance the same way: move the offending code tom2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.
Quoted 5 lines
Quoted 2 lines
That addresses this other thread, right?
https://www.mail-archive.com/[email protected]/msg720658.htmlCorrect! I wasn't aware of it, thanks for making the connection.
Thanks Markus for the fix.
Sundeep
Re: [PATCH for-5.1 1/2] msf2: Unbreak device-list-properties for "msf-soc"
Quoted 26 lines · sundeep subbaraya
Quoted 25 lines · Markus Armbruster
Quoted 22 lines · Philippe Mathieu-Daudé
Quoted 20 lines · Markus Armbruster
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio {"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}} {"execute": "qmp_capabilities"} {"return": {}} {"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}} Unsupported NIC model: ftgmac100 armbru@dusky:~/work/images$ echo $? 1This is what breaks "make check SPEED=slow".
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via
qemu_check_nic_model(). That's wrong.We fixed the exact same bug for device "allwinner-a10" in commit
8aabc5437b"hw/arm/allwinner-a10: Do not usend_tableininstance_initfunction". Fix this instance the same way: move the offending code tom2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.
Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
That addresses this other thread, right?
https://www.mail-archive.com/[email protected]/msg720658.htmlCorrect! I wasn't aware of it, thanks for making the connection.
Thanks Markus for the fix.
It's rather just a work-around that avoids the crash than a real fix. I think we can use it for the upcoming 5.1 release, but it would be great if you could rework this code for 5.2, so that the nd_table handling is moved to msf2-som.c instead.
Thanks,
Thomas
Re: [PATCH for-5.1 1/2] msf2: Unbreak device-list-properties for "msf-soc"
Quoted 33 lines · Thomas Huth
Quoted 32 lines · sundeep subbaraya
Quoted 30 lines · Markus Armbruster
Quoted 17 lines · Philippe Mathieu-Daudé
Quoted 16 lines · Markus Armbruster
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio {"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}} {"execute": "qmp_capabilities"} {"return": {}} {"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}} Unsupported NIC model: ftgmac100 armbru@dusky:~/work/images$ echo $? 1This is what breaks "make check SPEED=slow".
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via
qemu_check_nic_model(). That's wrong.Quoted 7 lines
Quoted 4 lines
We fixed the exact same bug for device "allwinner-a10" in commit
8aabc5437b"hw/arm/allwinner-a10: Do not usend_tableininstance_initfunction". Fix this instance the same way: move the offending code tom2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.That addresses this other thread, right?
https://www.mail-archive.com/[email protected]/msg720658.htmlCorrect! I wasn't aware of it, thanks for making the connection.
Quoted 6 lines
Quoted 1 line
Thanks Markus for the fix.
It's rather just a work-around that avoids the crash than a real fix. I think we can use it for the upcoming 5.1 release, but it would be great if you could rework this code for 5.2, so that the
nd_tablehandling is moved tomsf2-som.cinstead.
Yes this is not a 'fix' but rather a kludge.
Thomas gave some tips to work on a fix here:
https://www.mail-archive.com/[email protected]/msg720658.html
Thanks,
Phil.
Re: [PATCH for-5.1 1/2] msf2: Unbreak device-list-properties for "msf-soc"
Yep I will rework on this soon.
Thanks guys,
Sundeep
Quoted 49 lines · Philippe Mathieu-Daudé
Quoted 33 lines · Thomas Huth
Quoted 32 lines · sundeep subbaraya
Quoted 30 lines · Markus Armbruster
Quoted 17 lines · Philippe Mathieu-Daudé
Quoted 16 lines · Markus Armbruster
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio {"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}} {"execute": "qmp_capabilities"} {"return": {}} {"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}} Unsupported NIC model: ftgmac100 armbru@dusky:~/work/images$ echo $? 1This is what breaks "make check SPEED=slow".
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via
qemu_check_nic_model(). That's wrong.Quoted 7 lines
Quoted 4 lines
We fixed the exact same bug for device "allwinner-a10" in commit
8aabc5437b"hw/arm/allwinner-a10: Do not usend_tableininstance_initfunction". Fix this instance the same way: move the offending code tom2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.That addresses this other thread, right?
https://www.mail-archive.com/[email protected]/msg720658.htmlCorrect! I wasn't aware of it, thanks for making the connection.
Quoted 6 lines
Quoted 1 line
Thanks Markus for the fix.
It's rather just a work-around that avoids the crash than a real fix. I think we can use it for the upcoming 5.1 release, but it would be great if you could rework this code for 5.2, so that the
nd_tablehandling is moved tomsf2-som.cinstead.Yes this is not a 'fix' but rather a kludge.
Thomas gave some tips to work on a fix here:
https://www.mail-archive.com/[email protected]/msg720658.htmlThanks,
Phil.
Re: [PATCH for-5.1 1/2] msf2: Unbreak device-list-properties for "msf-soc"
In reply to Markus ArmbrusterQuoted 44 lines · Markus Armbruster
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio {"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}} {"execute": "qmp_capabilities"} {"return": {}} {"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}} Unsupported NIC model: ftgmac100 armbru@dusky:~/work/images$ echo $? 1This is what breaks "make check SPEED=slow".
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via
qemu_check_nic_model(). That's wrong.We fixed the exact same bug for device "allwinner-a10" in commit
8aabc5437b"hw/arm/allwinner-a10: Do not usend_tableininstance_initfunction". Fix this instance the same way: move the offending code tom2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.Fixes 05b7374a58cd18aa3516e33513808896d0ac9b7bSigned-off-by Markus Armbruster <[email protected]>--- hw/arm/msf2-soc.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)hw/arm/msf2-soc.c +0
-4@@ -82,10 +82,6 @@ static void m2sxxx_soc_initfn(Object *obj) }object_initialize_child (obj,"emac" , &s->emac,TYPE_MSS_EMAC ); -if (nd_table[0 ].used) { -qemu_check_nic_model (&nd_table[0 ],TYPE_MSS_EMAC ); -qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]); - } }static void m2sxxx_soc_realize (DeviceState *dev_soc,Error **errp)@@ -187,6 +183,11 @@ static void m2sxxx_soc_realize(DeviceState *dev_soc, Error **errp)g_free (bus_name);
hw/arm/msf2-soc.c static void m2sxxx_soc_realize(DeviceState *dev_soc, Error **errp) · line 193
Quoted 10 lines
hw/arm/msf2-soc.c +5
-0} +/* FIXME use qdev NIC properties instead of nd_table[] */ +if (nd_table[0 ].used) { +qemu_check_nic_model (&nd_table[0 ],TYPE_MSS_EMAC ); +qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]); + } dev =DEVICE (&s->emac);object_property_set_link (OBJECT (&s->emac),"ahb-bus" ,OBJECT (get_system_memory ()), &error_abort);
As long as nobody comes up with a proper clean up within the next days that moves the nd_table[] access to msf2-som.c, I think this is an acceptable work-around for 5.1.
Re: [PATCH for-5.1 1/2] msf2: Unbreak device-list-properties for "msf-soc"
In reply to Markus ArmbrusterQuoted 17 lines · Markus Armbruster
Watch this:
$ qemu-system-aarch64 -M ast2600-evb -S -display none -qmp stdio {"QMP": {"version": {"qemu": {"micro": 50, "minor": 0, "major": 5}, "package": "v5.0.0-2464-g3a9163af4e"}, "capabilities": ["oob"]}} {"execute": "qmp_capabilities"} {"return": {}} {"execute": "device-list-properties", "arguments": {"typename": "msf2-soc"}} Unsupported NIC model: ftgmac100 armbru@dusky:~/work/images$ echo $? 1This is what breaks "make check SPEED=slow".
Root cause is m2sxxx_soc_initfn()'s messing with nd_table[] via
qemu_check_nic_model(). That's wrong.
Quoted 7 lines
We fixed the exact same bug for device "allwinner-a10" in commit
8aabc5437b"hw/arm/allwinner-a10: Do not usend_tableininstance_initfunction". Fix this instance the same way: move the offending code tom2sxxx_soc_realize(), where it's less wrong, and add a FIXME comment.Fixes 05b7374a58cd18aa3516e33513808896d0ac9b7bSigned-off-by Markus Armbruster <[email protected]>
Alistair
Quoted 35 lines
--- hw/arm/msf2-soc.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)hw/arm/msf2-soc.c +5
-5@@ -82,10 +82,6 @@ static void m2sxxx_soc_initfn(Object *obj) }object_initialize_child (obj,"emac" , &s->emac,TYPE_MSS_EMAC ); -if (nd_table[0 ].used) { -qemu_check_nic_model (&nd_table[0 ],TYPE_MSS_EMAC ); -qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]); - } }static void m2sxxx_soc_realize (DeviceState *dev_soc,Error **errp)@@ -187,6 +183,11 @@ static void m2sxxx_soc_realize(DeviceState *dev_soc, Error **errp)g_free (bus_name); } +/* FIXME use qdev NIC properties instead of nd_table[] */ +if (nd_table[0 ].used) { +qemu_check_nic_model (&nd_table[0 ],TYPE_MSS_EMAC ); +qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]); + } dev =DEVICE (&s->emac);object_property_set_link (OBJECT (&s->emac),"ahb-bus" ,OBJECT (get_system_memory ()), &error_abort); --2.26.2
PATCH 2/2 hw: Mark nd_table[] misuse in realize methods FIXME
In reply to Markus Armbrusternd_table[] contains NIC configuration for boards to pick up. Device code has no business looking there. Several devices do it anyway. Two of them already have a suitable FIXME comment: "allwinner-a10" and "msf2-soc". Copy it to the others: "allwinner-h3", "xlnx-versal", "xlnx,zynqmp", "sparc32-ledma", "riscv.sifive.u.soc".
--- hw/arm/allwinner-h3.c | 1 + hw/arm/xlnx-versal.c | 1 + hw/arm/xlnx-zynqmp.c | 1 + hw/dma/sparc32_dma.c | 1 + hw/riscv/sifive_u.c | 1 + 5 files changed, 5 insertions(+)
hw/arm/allwinner-h3.c +1 -0
@@ -358,6 +358,7 @@ static void allwinner_h3_realize(DeviceState *dev, Error **errp) "sd-bus");/* EMAC */ +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd_table[0 ].used) {qemu_check_nic_model (&nd_table[0 ],TYPE_AW_SUN8I_EMAC );qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]);
hw/arm/xlnx-versal.c +1 -0
@@ -160,6 +160,7 @@ static void versal_create_gems(Versal *s, qemu_irq *pic)object_initialize_child (OBJECT (s), name, &s->lpd.iou.gem[i],TYPE_CADENCE_GEM ); dev =DEVICE (&s->lpd.iou.gem[i]); +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model (nd,"cadence_gem" );qdev_set_nic_properties (dev, nd);
hw/arm/xlnx-zynqmp.c +1 -0
@@ -455,6 +455,7 @@ static void xlnx_zynqmp_realize(DeviceState *dev, Error **errp)for (i =0 ; i <XLNX_ZYNQMP_NUM_GEMS ; i++) {NICInfo *nd = &nd_table[i]; +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model (nd,TYPE_CADENCE_GEM );qdev_set_nic_properties (DEVICE (&s->gem[i]), nd);
hw/dma/sparc32_dma.c +1 -0
@@ -341,6 +341,7 @@ static void sparc32_ledma_device_realize(DeviceState *dev, Error **errp)DeviceState *d;NICInfo *nd = &nd_table[0 ]; +/* FIXME use qdev NIC properties instead of nd_table[] */ qemu_check_nic_model (nd,TYPE_LANCE ); d =qdev_new (TYPE_LANCE );
hw/riscv/sifive_u.c +1 -0
@@ -714,6 +714,7 @@ static void sifive_u_soc_realize(DeviceState *dev, Error **errp) }sysbus_mmio_map (SYS_BUS_DEVICE (&s->otp),0 , memmap[SIFIVE_U_OTP ].base); +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model (nd,TYPE_CADENCE_GEM );qdev_set_nic_properties (DEVICE (&s->gem), nd);
Signature · 1 line
2.26.2
Re: [PATCH for-5.1 2/2] hw: Mark nd_table[] misuse in realize methods FIXME
Quoted 6 lines · Markus Armbruster
nd_table[] contains NIC configuration for boards to pick up. Device code has no business looking there. Several devices do it anyway. Two of them already have a suitable FIXME comment: "allwinner-a10" and "msf2-soc". Copy it to the others: "allwinner-h3", "xlnx-versal", "xlnx,zynqmp", "sparc32-ledma", "riscv.sifive.u.soc".
Quoted 8 lines
Signed-off-by Markus Armbruster <[email protected]>--- hw/arm/allwinner-h3.c | 1 + hw/arm/xlnx-versal.c | 1 + hw/arm/xlnx-zynqmp.c | 1 + hw/dma/sparc32_dma.c | 1 + hw/riscv/sifive_u.c | 1 + 5 files changed, 5 insertions(+)
Maybe we should add an entry to
https://wiki.qemu.org/Contribute/BiteSizedTasks for this?
Re: [PATCH for-5.1 2/2] hw: Mark nd_table[] misuse in realize methods FIXME
Quoted 16 lines · Thomas Huth
Quoted 6 lines · Markus Armbruster
nd_table[] contains NIC configuration for boards to pick up. Device code has no business looking there. Several devices do it anyway. Two of them already have a suitable FIXME comment: "allwinner-a10" and "msf2-soc". Copy it to the others: "allwinner-h3", "xlnx-versal", "xlnx,zynqmp", "sparc32-ledma", "riscv.sifive.u.soc".
Quoted 8 lines
Signed-off-by Markus Armbruster <[email protected]>--- hw/arm/allwinner-h3.c | 1 + hw/arm/xlnx-versal.c | 1 + hw/arm/xlnx-zynqmp.c | 1 + hw/dma/sparc32_dma.c | 1 + hw/riscv/sifive_u.c | 1 + 5 files changed, 5 insertions(+)
Quoted 2 lines
Maybe we should add an entry to
https://wiki.qemu.org/Contribute/BiteSizedTasks for this?
I'd prefer to leave that to someone who knows how this should be done properly, and thus has a better idea on how big the bite actually is. Figuring this out myself might be more work than fixing it.
Quoted 1 line
Reviewed-by Thomas Huth <[email protected]>
Thanks!
Re: [PATCH for-5.1 2/2] hw: Mark nd_table[] misuse in realize methods FIXME
In reply to Markus ArmbrusterQuoted 8 lines · Markus Armbruster
nd_table[] contains NIC configuration for boards to pick up. Device code has no business looking there. Several devices do it anyway. Two of them already have a suitable FIXME comment: "allwinner-a10" and "msf2-soc". Copy it to the others: "allwinner-h3", "xlnx-versal", "xlnx,zynqmp", "sparc32-ledma", "riscv.sifive.u.soc".
Signed-off-by Markus Armbruster <[email protected]>
Alistair
Quoted 72 lines
--- hw/arm/allwinner-h3.c | 1 + hw/arm/xlnx-versal.c | 1 + hw/arm/xlnx-zynqmp.c | 1 + hw/dma/sparc32_dma.c | 1 + hw/riscv/sifive_u.c | 1 + 5 files changed, 5 insertions(+)hw/arm/allwinner-h3.c +1
-0@@ -358,6 +358,7 @@ static void allwinner_h3_realize(DeviceState *dev, Error **errp) "sd-bus");/* EMAC */ +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd_table[0 ].used) {qemu_check_nic_model (&nd_table[0 ],TYPE_AW_SUN8I_EMAC );qdev_set_nic_properties (DEVICE (&s->emac), &nd_table[0 ]);hw/arm/xlnx-versal.c +1
-0@@ -160,6 +160,7 @@ static void versal_create_gems(Versal *s, qemu_irq *pic)object_initialize_child (OBJECT (s), name, &s->lpd.iou.gem[i],TYPE_CADENCE_GEM ); dev =DEVICE (&s->lpd.iou.gem[i]); +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model (nd,"cadence_gem" );qdev_set_nic_properties (dev, nd);hw/arm/xlnx-zynqmp.c +1
-0@@ -455,6 +455,7 @@ static void xlnx_zynqmp_realize(DeviceState *dev, Error **errp)for (i =0 ; i <XLNX_ZYNQMP_NUM_GEMS ; i++) {NICInfo *nd = &nd_table[i]; +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model (nd,TYPE_CADENCE_GEM );qdev_set_nic_properties (DEVICE (&s->gem[i]), nd);hw/dma/sparc32_dma.c +1
-0@@ -341,6 +341,7 @@ static void sparc32_ledma_device_realize(DeviceState *dev, Error **errp)DeviceState *d;NICInfo *nd = &nd_table[0 ]; +/* FIXME use qdev NIC properties instead of nd_table[] */ qemu_check_nic_model (nd,TYPE_LANCE ); d =qdev_new (TYPE_LANCE );hw/riscv/sifive_u.c +1
-1@@ -714,6 +714,7 @@ static void sifive_u_soc_realize(DeviceState *dev, Error **errp) }sysbus_mmio_map (SYS_BUS_DEVICE (&s->otp),0 , memmap[SIFIVE_U_OTP ].base); +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model (nd,TYPE_CADENCE_GEM );qdev_set_nic_properties (DEVICE (&s->gem), nd); --2.26.2
Re: [PATCH for-5.1 2/2] hw: Mark nd_table[] misuse in realize methods FIXME
In reply to Markus ArmbrusterQuoted 8 lines · On Wed, Jul 15, 2020, 16:04 Markus Armbruster <[email protected]> wrote:
nd_table[] contains NIC configuration for boards to pick up. Device code has no business looking there. Several devices do it anyway. Two of them already have a suitable FIXME comment: "allwinner-a10" and "msf2-soc". Copy it to the others: "allwinner-h3", "xlnx-versal", "xlnx,zynqmp", "sparc32-ledma", "riscv.sifive.u.soc".
Signed-off-by Markus Armbruster <[email protected]>
---
Quoted 77 lines
hw/arm/allwinner-h3.c | 1 + hw/arm/xlnx-versal.c | 1 + hw/arm/xlnx-zynqmp.c | 1 + hw/dma/sparc32_dma.c | 1 + hw/riscv/sifive_u.c | 1 + 5 files changed, 5 insertions(+)hw/arm/allwinner-h3.c +1
-0@@ -358,6 +358,7 @@ static void allwinner_h3_realize(DeviceState *dev, Error **errp) "sd-bus");/* EMAC */ +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd_table[0 ].used) {qemu_check_nic_model (&nd_table[0 ],TYPE_AW_SUN8I_EMAC );qdev_set_nic_properties(DEVICE(&s->emac), &nd_table[0]);hw/arm/xlnx-versal.c +1
-0@@ -160,6 +160,7 @@ static void versal_create_gems(Versal *s, qemu_irq *pic)object_initialize_child (OBJECT (s), name, &s->lpd.iou.gem[i],TYPE_CADENCE_GEM ); dev =DEVICE (&s->lpd.iou.gem[i]); +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model (nd,"cadence_gem" );qdev_set_nic_properties(dev, nd);hw/arm/xlnx-zynqmp.c +1
-0@@ -455,6 +455,7 @@ static void xlnx_zynqmp_realize(DeviceState *dev, Error **errp)for (i =0 ; i <XLNX_ZYNQMP_NUM_GEMS ; i++) {NICInfo *nd = &nd_table[i]; +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model (nd,TYPE_CADENCE_GEM );qdev_set_nic_properties(DEVICE(&s->gem[i]), nd);hw/dma/sparc32_dma.c +1
-0@@ -341,6 +341,7 @@ static void sparc32_ledma_device_realize(DeviceState *dev, Error **errp)DeviceState *d;NICInfo *nd = &nd_table[0 ]; +/* FIXME use qdev NIC properties instead of nd_table[] */ qemu_check_nic_model (nd,TYPE_LANCE );d = qdev_new(TYPE_LANCE);hw/riscv/sifive_u.c +1
-0@@ -714,6 +714,7 @@ static void sifive_u_soc_realize(DeviceState *dev, Error **errp) }sysbus_mmio_map (SYS_BUS_DEVICE (&s->otp),0 , memmap[SIFIVE_U_OTP ].base); +/* FIXME use qdev NIC properties instead of nd_table[] */ if (nd->used) {qemu_check_nic_model(nd, TYPE_CADENCE_GEM); qdev_set_nic_properties(DEVICE(&s->gem), nd);--
2.26.2
Patchew URL: https://patchew.org/QEMU/[email protected]/
Patchew URL: https://patchew.org/QEMU/[email protected]/
Hi,
This series failed the docker-quick@centos7 build test. Please find the testing commands and their output below. If you have Docker installed, you can probably reproduce it locally.
=== TEST SCRIPT BEGIN ===
#!/bin/bash
make docker-image-centos7 V=1 NETWORK=1
time make docker-test-quick@centos7 SHOW_ENV=1 J=14 NETWORK=1 === TEST SCRIPT END ===
qemu-system-aarch64: falling back to tcg
TEST check-unit: tests/test-char
**
ERROR:/tmp/qemu-test/src/tests/test-char.c:1204:char_serial_test: 'chr' should not be NULL
ERROR test-char - Bail out! ERROR:/tmp/qemu-test/src/tests/test-char.c:1204:char_serial_test: 'chr' should not be NULL
make: *** [check-unit] Error 1
make: *** Waiting for unfinished jobs....
TEST iotest-qcow2: 025
TEST iotest-qcow2: 027
---
raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['sudo', '-n', 'docker', 'run', '--label', 'com.qemu.instance.uuid=aa283144540a4d70bf303fb95ee05619', '-u', '1003', '--security-opt', 'seccomp=unconfined', '--rm', '-e', 'TARGET_LIST=', '-e', 'EXTRA_CONFIGURE_OPTS=', '-e', 'V=', '-e', 'J=14', '-e', 'DEBUG=', '-e', 'SHOW_ENV=1', '-e', 'CCACHE_DIR=/var/tmp/ccache', '-v', '/home/patchew2/.cache/qemu-docker-ccache:/var/tmp/ccache:z', '-v', '/var/tmp/patchew-tester-tmp-sn6v_tr1/src/docker-src.2020-07-15-10.15.38.17277:/var/tmp/qemu:z,ro', 'qemu/centos7', '/var/tmp/qemu/run', 'test-quick']' returned non-zero exit status 2. filter=--filter=label=com.qemu.instance.uuid=aa283144540a4d70bf303fb95ee05619
make[1]: *** [docker-run] Error 1
make[1]: Leaving directory `/var/tmp/patchew-tester-tmp-sn6v_tr1/src'
make: *** [docker-run-test-quick@centos7] Error 2
real 14m52.787s
user 0m8.650s
The full log is available at
http://patchew.org/logs/[email protected]/testing.docker-quick@centos7/?type=message.
---
Email generated automatically by Patchew [https://patchew.org/].
Please send your feedback to [email protected]