Message ID | 1506082711-16004-1-git-send-email-mark.cave-ayland@ilande.co.uk |
---|---|
State | New |
Headers | show |
Series | pci: allow 32-bit PCI IO accesses to pass through the PCI bridge | expand |
On 09/22/2017 07:18 AM, Mark Cave-Ayland wrote: > Whilst the underlying PCI bridge implementation supports 32-bit PCI IO > accesses, unfortunately they are truncated at the legacy 64K limit. > > Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> > --- > hw/pci/pci_bridge.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) Reviewed-by: Richard Henderson <richard.henderson@linaro.org> r~
On 09/22/17 14:18, Mark Cave-Ayland wrote: > Whilst the underlying PCI bridge implementation supports 32-bit PCI IO > accesses, unfortunately they are truncated at the legacy 64K limit. > > Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> > --- > hw/pci/pci_bridge.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c > index 17feae5..a47d257 100644 > --- a/hw/pci/pci_bridge.c > +++ b/hw/pci/pci_bridge.c > @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) > sec_bus->address_space_mem = &br->address_space_mem; > memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); > sec_bus->address_space_io = &br->address_space_io; > - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); > + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", > + UINT32_MAX); > br->windows = pci_bridge_region_init(br); > QLIST_INIT(&sec_bus->child); > QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); > Based on the commit message, I assume this change is guest-visible. If so, should it be made dependent on a compat property, so that it doesn't cause problems with migration? Thanks, Laszlo
On 22/09/17 23:18, Laszlo Ersek wrote: > On 09/22/17 14:18, Mark Cave-Ayland wrote: >> Whilst the underlying PCI bridge implementation supports 32-bit PCI IO >> accesses, unfortunately they are truncated at the legacy 64K limit. >> >> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> >> --- >> hw/pci/pci_bridge.c | 3 ++- >> 1 file changed, 2 insertions(+), 1 deletion(-) >> >> diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c >> index 17feae5..a47d257 100644 >> --- a/hw/pci/pci_bridge.c >> +++ b/hw/pci/pci_bridge.c >> @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) >> sec_bus->address_space_mem = &br->address_space_mem; >> memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); >> sec_bus->address_space_io = &br->address_space_io; >> - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); >> + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", >> + UINT32_MAX); >> br->windows = pci_bridge_region_init(br); >> QLIST_INIT(&sec_bus->child); >> QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); >> > > Based on the commit message, I assume this change is guest-visible. If > so, should it be made dependent on a compat property, so that it doesn't > cause problems with migration? In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, so unless a PCI bridge has this bit set then it's impossible for this change to be guest visible. I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users (other than an upcoming patchset from me!), so this combined with the fact that without this patch the feature is broken makes me think that I am the first user and so existing guests won't have a problem. ATB, Mark.
On 23/09/2017 11:23, Mark Cave-Ayland wrote: > On 22/09/17 23:18, Laszlo Ersek wrote: > >> On 09/22/17 14:18, Mark Cave-Ayland wrote: >>> Whilst the underlying PCI bridge implementation supports 32-bit PCI IO >>> accesses, unfortunately they are truncated at the legacy 64K limit. >>> >>> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> >>> --- >>> hw/pci/pci_bridge.c | 3 ++- >>> 1 file changed, 2 insertions(+), 1 deletion(-) >>> >>> diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c >>> index 17feae5..a47d257 100644 >>> --- a/hw/pci/pci_bridge.c >>> +++ b/hw/pci/pci_bridge.c >>> @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) >>> sec_bus->address_space_mem = &br->address_space_mem; >>> memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); >>> sec_bus->address_space_io = &br->address_space_io; >>> - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); >>> + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", >>> + UINT32_MAX); >>> br->windows = pci_bridge_region_init(br); >>> QLIST_INIT(&sec_bus->child); >>> QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); >>> >> Hi Mark, >> Based on the commit message, I assume this change is guest-visible. If >> so, should it be made dependent on a compat property, so that it doesn't >> cause problems with migration? > > In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 > in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, > so unless a PCI bridge has this bit set then it's impossible for this > change to be guest visible. > > I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users > (other than an upcoming patchset from me!), so this combined with the > fact that without this patch the feature is broken makes me think that I > am the first user and so existing guests won't have a problem. > (adding Dave for his expertise) Do you know how the migration code will behave if it will have a 65k address space on source and MAX UINT on destination? (and the other way around for rolling back) Thanks, Marcel > > ATB, > > Mark. >
On 24/09/17 16:43, Marcel Apfelbaum wrote: > Hi Mark, > >>> Based on the commit message, I assume this change is guest-visible. If >>> so, should it be made dependent on a compat property, so that it doesn't >>> cause problems with migration? >> >> In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 >> in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, >> so unless a PCI bridge has this bit set then it's impossible for this >> change to be guest visible. >> >> I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users >> (other than an upcoming patchset from me!), so this combined with the >> fact that without this patch the feature is broken makes me think that I >> am the first user and so existing guests won't have a problem. >> > > (adding Dave for his expertise) > > Do you know how the migration code will behave if it will have > a 65k address space on source and MAX UINT on destination? > (and the other way around for rolling back) Thanks Marcel. I should add that qemu-system-sparc64 isn't currently migratable anyhow, so if with my upcoming patch qemu-system-sparc64 is still the only user of PCI_IO_RANGE_TYPE_32 then that won't cause me any particular issue trying to migrate to earlier versions. Also in my local tests without the patch applied, the guest always panics during boot trying to access the IO space above 64K so I can't see there's a way that an older guest could boot in order to migrate forward either. ATB, Mark.
* Marcel Apfelbaum (marcel@redhat.com) wrote: > On 23/09/2017 11:23, Mark Cave-Ayland wrote: > > On 22/09/17 23:18, Laszlo Ersek wrote: > > > > > On 09/22/17 14:18, Mark Cave-Ayland wrote: > > > > Whilst the underlying PCI bridge implementation supports 32-bit PCI IO > > > > accesses, unfortunately they are truncated at the legacy 64K limit. > > > > > > > > Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> > > > > --- > > > > hw/pci/pci_bridge.c | 3 ++- > > > > 1 file changed, 2 insertions(+), 1 deletion(-) > > > > > > > > diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c > > > > index 17feae5..a47d257 100644 > > > > --- a/hw/pci/pci_bridge.c > > > > +++ b/hw/pci/pci_bridge.c > > > > @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) > > > > sec_bus->address_space_mem = &br->address_space_mem; > > > > memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); > > > > sec_bus->address_space_io = &br->address_space_io; > > > > - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); > > > > + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", > > > > + UINT32_MAX); > > > > br->windows = pci_bridge_region_init(br); > > > > QLIST_INIT(&sec_bus->child); > > > > QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); > > > > > > > > > Hi Mark, > > > > Based on the commit message, I assume this change is guest-visible. If > > > so, should it be made dependent on a compat property, so that it doesn't > > > cause problems with migration? > > > > In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 > > in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, > > so unless a PCI bridge has this bit set then it's impossible for this > > change to be guest visible. > > > > I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users > > (other than an upcoming patchset from me!), so this combined with the > > fact that without this patch the feature is broken makes me think that I > > am the first user and so existing guests won't have a problem. > > > > (adding Dave for his expertise) > > Do you know how the migration code will behave if it will have > a 65k address space on source and MAX UINT on destination? > (and the other way around for rolling back) Hmm not sure; we don't migrate regions explicitly; just RAMBlocks and devices that back them. If the change is visible in the IO addresses allocated to the PCI devices or in the config space then it might fail. Dave > Thanks, > Marcel > > > > > ATB, > > > > Mark. > > > -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
On 25/09/17 09:11, Dr. David Alan Gilbert wrote: > * Marcel Apfelbaum (marcel@redhat.com) wrote: >> On 23/09/2017 11:23, Mark Cave-Ayland wrote: >>> On 22/09/17 23:18, Laszlo Ersek wrote: >>> >>>> On 09/22/17 14:18, Mark Cave-Ayland wrote: >>>>> Whilst the underlying PCI bridge implementation supports 32-bit PCI IO >>>>> accesses, unfortunately they are truncated at the legacy 64K limit. >>>>> >>>>> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> >>>>> --- >>>>> hw/pci/pci_bridge.c | 3 ++- >>>>> 1 file changed, 2 insertions(+), 1 deletion(-) >>>>> >>>>> diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c >>>>> index 17feae5..a47d257 100644 >>>>> --- a/hw/pci/pci_bridge.c >>>>> +++ b/hw/pci/pci_bridge.c >>>>> @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) >>>>> sec_bus->address_space_mem = &br->address_space_mem; >>>>> memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); >>>>> sec_bus->address_space_io = &br->address_space_io; >>>>> - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); >>>>> + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", >>>>> + UINT32_MAX); >>>>> br->windows = pci_bridge_region_init(br); >>>>> QLIST_INIT(&sec_bus->child); >>>>> QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); >>>>> >>>> >> >> Hi Mark, >> >>>> Based on the commit message, I assume this change is guest-visible. If >>>> so, should it be made dependent on a compat property, so that it doesn't >>>> cause problems with migration? >>> >>> In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 >>> in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, >>> so unless a PCI bridge has this bit set then it's impossible for this >>> change to be guest visible. >>> >>> I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users >>> (other than an upcoming patchset from me!), so this combined with the >>> fact that without this patch the feature is broken makes me think that I >>> am the first user and so existing guests won't have a problem. >>> >> >> (adding Dave for his expertise) >> >> Do you know how the migration code will behave if it will have >> a 65k address space on source and MAX UINT on destination? >> (and the other way around for rolling back) > > Hmm not sure; we don't migrate regions explicitly; just RAMBlocks > and devices that back them. If the change is visible in the IO > addresses allocated to the PCI devices or in the config space then > it might fail. For reference here is the link to the sun4u patch I posted yesterday that requires this fix if anyone else would like to test: https://lists.gnu.org/archive/html/qemu-devel/2017-09/msg07355.html. Other than that are there any further objections to this patch? ATB, Mark.
On 09/28/17 09:31, Mark Cave-Ayland wrote: > On 25/09/17 09:11, Dr. David Alan Gilbert wrote: > >> * Marcel Apfelbaum (marcel@redhat.com) wrote: >>> On 23/09/2017 11:23, Mark Cave-Ayland wrote: >>>> On 22/09/17 23:18, Laszlo Ersek wrote: >>>> >>>>> On 09/22/17 14:18, Mark Cave-Ayland wrote: >>>>>> Whilst the underlying PCI bridge implementation supports 32-bit PCI IO >>>>>> accesses, unfortunately they are truncated at the legacy 64K limit. >>>>>> >>>>>> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> >>>>>> --- >>>>>> hw/pci/pci_bridge.c | 3 ++- >>>>>> 1 file changed, 2 insertions(+), 1 deletion(-) >>>>>> >>>>>> diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c >>>>>> index 17feae5..a47d257 100644 >>>>>> --- a/hw/pci/pci_bridge.c >>>>>> +++ b/hw/pci/pci_bridge.c >>>>>> @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) >>>>>> sec_bus->address_space_mem = &br->address_space_mem; >>>>>> memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); >>>>>> sec_bus->address_space_io = &br->address_space_io; >>>>>> - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); >>>>>> + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", >>>>>> + UINT32_MAX); >>>>>> br->windows = pci_bridge_region_init(br); >>>>>> QLIST_INIT(&sec_bus->child); >>>>>> QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); >>>>>> >>>>> >>> >>> Hi Mark, >>> >>>>> Based on the commit message, I assume this change is guest-visible. If >>>>> so, should it be made dependent on a compat property, so that it doesn't >>>>> cause problems with migration? >>>> >>>> In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 >>>> in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, >>>> so unless a PCI bridge has this bit set then it's impossible for this >>>> change to be guest visible. >>>> >>>> I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users >>>> (other than an upcoming patchset from me!), so this combined with the >>>> fact that without this patch the feature is broken makes me think that I >>>> am the first user and so existing guests won't have a problem. >>>> >>> >>> (adding Dave for his expertise) >>> >>> Do you know how the migration code will behave if it will have >>> a 65k address space on source and MAX UINT on destination? >>> (and the other way around for rolling back) >> >> Hmm not sure; we don't migrate regions explicitly; just RAMBlocks >> and devices that back them. If the change is visible in the IO >> addresses allocated to the PCI devices or in the config space then >> it might fail. > > For reference here is the link to the sun4u patch I posted yesterday > that requires this fix if anyone else would like to test: > https://lists.gnu.org/archive/html/qemu-devel/2017-09/msg07355.html. > > Other than that are there any further objections to this patch? None from my side. (I didn't "object" to begin with :) , I was just curious about any possible migration impact.) Thanks! Laszlo
On 28/09/2017 10:31, Mark Cave-Ayland wrote: > On 25/09/17 09:11, Dr. David Alan Gilbert wrote: > >> * Marcel Apfelbaum (marcel@redhat.com) wrote: >>> On 23/09/2017 11:23, Mark Cave-Ayland wrote: >>>> On 22/09/17 23:18, Laszlo Ersek wrote: >>>> >>>>> On 09/22/17 14:18, Mark Cave-Ayland wrote: >>>>>> Whilst the underlying PCI bridge implementation supports 32-bit PCI IO >>>>>> accesses, unfortunately they are truncated at the legacy 64K limit. >>>>>> >>>>>> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> >>>>>> --- >>>>>> hw/pci/pci_bridge.c | 3 ++- >>>>>> 1 file changed, 2 insertions(+), 1 deletion(-) >>>>>> >>>>>> diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c >>>>>> index 17feae5..a47d257 100644 >>>>>> --- a/hw/pci/pci_bridge.c >>>>>> +++ b/hw/pci/pci_bridge.c >>>>>> @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) >>>>>> sec_bus->address_space_mem = &br->address_space_mem; >>>>>> memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); >>>>>> sec_bus->address_space_io = &br->address_space_io; >>>>>> - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); >>>>>> + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", >>>>>> + UINT32_MAX); >>>>>> br->windows = pci_bridge_region_init(br); >>>>>> QLIST_INIT(&sec_bus->child); >>>>>> QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); >>>>>> >>>>> >>> >>> Hi Mark, >>> >>>>> Based on the commit message, I assume this change is guest-visible. If >>>>> so, should it be made dependent on a compat property, so that it doesn't >>>>> cause problems with migration? >>>> >>>> In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 >>>> in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, >>>> so unless a PCI bridge has this bit set then it's impossible for this >>>> change to be guest visible. >>>> >>>> I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users >>>> (other than an upcoming patchset from me!), so this combined with the >>>> fact that without this patch the feature is broken makes me think that I >>>> am the first user and so existing guests won't have a problem. >>>> >>> >>> (adding Dave for his expertise) >>> >>> Do you know how the migration code will behave if it will have >>> a 65k address space on source and MAX UINT on destination? >>> (and the other way around for rolling back) >> >> Hmm not sure; we don't migrate regions explicitly; just RAMBlocks >> and devices that back them. If the change is visible in the IO >> addresses allocated to the PCI devices or in the config space then >> it might fail. > > For reference here is the link to the sun4u patch I posted yesterday > that requires this fix if anyone else would like to test: > https://lists.gnu.org/archive/html/qemu-devel/2017-09/msg07355.html. > > Other than that are there any further objections to this patch? > Since we don't support 32-bit IO accesses (yet) in x86 and the address space is not visible to guest, it should be OK. I'll run a simple migration test to be sure, I'll let you know if something goes wrong. Thanks, Marcel > > ATB, > > Mark. >
On 01/10/17 22:44, Mark Cave-Ayland wrote: > On 28/09/17 08:56, Laszlo Ersek wrote: > >> On 09/28/17 09:31, Mark Cave-Ayland wrote: >>> On 25/09/17 09:11, Dr. David Alan Gilbert wrote: >>> >>>> * Marcel Apfelbaum (marcel@redhat.com) wrote: >>>>> On 23/09/2017 11:23, Mark Cave-Ayland wrote: >>>>>> On 22/09/17 23:18, Laszlo Ersek wrote: >>>>>> >>>>>>> On 09/22/17 14:18, Mark Cave-Ayland wrote: >>>>>>>> Whilst the underlying PCI bridge implementation supports 32-bit PCI IO >>>>>>>> accesses, unfortunately they are truncated at the legacy 64K limit. >>>>>>>> >>>>>>>> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> >>>>>>>> --- >>>>>>>> hw/pci/pci_bridge.c | 3 ++- >>>>>>>> 1 file changed, 2 insertions(+), 1 deletion(-) >>>>>>>> >>>>>>>> diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c >>>>>>>> index 17feae5..a47d257 100644 >>>>>>>> --- a/hw/pci/pci_bridge.c >>>>>>>> +++ b/hw/pci/pci_bridge.c >>>>>>>> @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) >>>>>>>> sec_bus->address_space_mem = &br->address_space_mem; >>>>>>>> memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); >>>>>>>> sec_bus->address_space_io = &br->address_space_io; >>>>>>>> - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); >>>>>>>> + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", >>>>>>>> + UINT32_MAX); >>>>>>>> br->windows = pci_bridge_region_init(br); >>>>>>>> QLIST_INIT(&sec_bus->child); >>>>>>>> QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); >>>>>>>> >>>>>>> >>>>> >>>>> Hi Mark, >>>>> >>>>>>> Based on the commit message, I assume this change is guest-visible. If >>>>>>> so, should it be made dependent on a compat property, so that it doesn't >>>>>>> cause problems with migration? >>>>>> >>>>>> In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 >>>>>> in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, >>>>>> so unless a PCI bridge has this bit set then it's impossible for this >>>>>> change to be guest visible. >>>>>> >>>>>> I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users >>>>>> (other than an upcoming patchset from me!), so this combined with the >>>>>> fact that without this patch the feature is broken makes me think that I >>>>>> am the first user and so existing guests won't have a problem. >>>>>> >>>>> >>>>> (adding Dave for his expertise) >>>>> >>>>> Do you know how the migration code will behave if it will have >>>>> a 65k address space on source and MAX UINT on destination? >>>>> (and the other way around for rolling back) >>>> >>>> Hmm not sure; we don't migrate regions explicitly; just RAMBlocks >>>> and devices that back them. If the change is visible in the IO >>>> addresses allocated to the PCI devices or in the config space then >>>> it might fail. >>> >>> For reference here is the link to the sun4u patch I posted yesterday >>> that requires this fix if anyone else would like to test: >>> https://lists.gnu.org/archive/html/qemu-devel/2017-09/msg07355.html. >>> >>> Other than that are there any further objections to this patch? >> >> None from my side. (I didn't "object" to begin with :) , I was just >> curious about any possible migration impact.) > > Okay great! I guess I mis-read your query as being a NACK for the patch. Hi Marcel, Did you manage to get any further with your testing? It appears that freeze is coming soon and my remaining changes for the sun4u machine are currently blocked on this patch :/ ATB, Mark.
On 11/10/2017 9:35, Mark Cave-Ayland wrote: > On 01/10/17 22:44, Mark Cave-Ayland wrote: > >> On 28/09/17 08:56, Laszlo Ersek wrote: >> >>> On 09/28/17 09:31, Mark Cave-Ayland wrote: >>>> On 25/09/17 09:11, Dr. David Alan Gilbert wrote: >>>> >>>>> * Marcel Apfelbaum (marcel@redhat.com) wrote: >>>>>> On 23/09/2017 11:23, Mark Cave-Ayland wrote: >>>>>>> On 22/09/17 23:18, Laszlo Ersek wrote: >>>>>>> >>>>>>>> On 09/22/17 14:18, Mark Cave-Ayland wrote: >>>>>>>>> Whilst the underlying PCI bridge implementation supports 32-bit PCI IO >>>>>>>>> accesses, unfortunately they are truncated at the legacy 64K limit. >>>>>>>>> >>>>>>>>> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> >>>>>>>>> --- >>>>>>>>> hw/pci/pci_bridge.c | 3 ++- >>>>>>>>> 1 file changed, 2 insertions(+), 1 deletion(-) >>>>>>>>> >>>>>>>>> diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c >>>>>>>>> index 17feae5..a47d257 100644 >>>>>>>>> --- a/hw/pci/pci_bridge.c >>>>>>>>> +++ b/hw/pci/pci_bridge.c >>>>>>>>> @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) >>>>>>>>> sec_bus->address_space_mem = &br->address_space_mem; >>>>>>>>> memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); >>>>>>>>> sec_bus->address_space_io = &br->address_space_io; >>>>>>>>> - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); >>>>>>>>> + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", >>>>>>>>> + UINT32_MAX); >>>>>>>>> br->windows = pci_bridge_region_init(br); >>>>>>>>> QLIST_INIT(&sec_bus->child); >>>>>>>>> QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling); >>>>>>>>> >>>>>>>> >>>>>> >>>>>> Hi Mark, >>>>>> >>>>>>>> Based on the commit message, I assume this change is guest-visible. If >>>>>>>> so, should it be made dependent on a compat property, so that it doesn't >>>>>>>> cause problems with migration? >>>>>>> >>>>>>> In order to enable 32-bit IO accesses the PCI bridge needs to set bit 0 >>>>>>> in the IO_LIMIT and IO_BASE registers - this bit is read-only to guests, >>>>>>> so unless a PCI bridge has this bit set then it's impossible for this >>>>>>> change to be guest visible. >>>>>>> >>>>>>> I did a grep for PCI_IO_RANGE_TYPE_32 and didn't see any existing users >>>>>>> (other than an upcoming patchset from me!), so this combined with the >>>>>>> fact that without this patch the feature is broken makes me think that I >>>>>>> am the first user and so existing guests won't have a problem. >>>>>>> >>>>>> >>>>>> (adding Dave for his expertise) >>>>>> >>>>>> Do you know how the migration code will behave if it will have >>>>>> a 65k address space on source and MAX UINT on destination? >>>>>> (and the other way around for rolling back) >>>>> >>>>> Hmm not sure; we don't migrate regions explicitly; just RAMBlocks >>>>> and devices that back them. If the change is visible in the IO >>>>> addresses allocated to the PCI devices or in the config space then >>>>> it might fail. >>>> >>>> For reference here is the link to the sun4u patch I posted yesterday >>>> that requires this fix if anyone else would like to test: >>>> https://lists.gnu.org/archive/html/qemu-devel/2017-09/msg07355.html. >>>> >>>> Other than that are there any further objections to this patch? >>> >>> None from my side. (I didn't "object" to begin with :) , I was just >>> curious about any possible migration impact.) >> >> Okay great! I guess I mis-read your query as being a NACK for the patch. > > Hi Marcel, > > Did you manage to get any further with your testing? It appears that > freeze is coming soon and my remaining changes for the sun4u machine are > currently blocked on this patch :/ > Not yet, but I see your patch is already merged. I'll let you know if there will be any problems. Thanks, Marcel > > ATB, > > Mark. >
diff --git a/hw/pci/pci_bridge.c b/hw/pci/pci_bridge.c index 17feae5..a47d257 100644 --- a/hw/pci/pci_bridge.c +++ b/hw/pci/pci_bridge.c @@ -379,7 +379,8 @@ void pci_bridge_initfn(PCIDevice *dev, const char *typename) sec_bus->address_space_mem = &br->address_space_mem; memory_region_init(&br->address_space_mem, OBJECT(br), "pci_bridge_pci", UINT64_MAX); sec_bus->address_space_io = &br->address_space_io; - memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", 65536); + memory_region_init(&br->address_space_io, OBJECT(br), "pci_bridge_io", + UINT32_MAX); br->windows = pci_bridge_region_init(br); QLIST_INIT(&sec_bus->child); QLIST_INSERT_HEAD(&parent->child, sec_bus, sibling);
Whilst the underlying PCI bridge implementation supports 32-bit PCI IO accesses, unfortunately they are truncated at the legacy 64K limit. Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> --- hw/pci/pci_bridge.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)