[Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Michael S. Tsirkin]

Exactly the same thing as v1 but rebased to master to
help Peter resolve conflicts, so I'm not bothering
with resending all patches.

A huge patchset, but the scariest part is Igor's patches,
and these have been used by multiple people by now.
virtio header change is a bit rushed, but getting them upstream
seems like the best way to give them the cross-platform
testing that they need, and this takes us a step closer to
virtio-1.0 support.

The following changes since commit 11d39a131020cc5c54ff9bc86d3259f7d32bf849:

  Merge remote-tracking branch 'remotes/borntraeger/tags/s390x-20150218' into staging (2015-02-26 09:08:54 +0000)

are available in the git repository at:

  git://git.kernel.org/pub/scm/virt/kvm/mst/qemu.git tags/for_upstream_rebased

for you to fetch changes up to dc33c0293505db9c747a6e27487a808ab2b62a6f:

  acpi-test: update expected files (2015-02-26 13:04:19 +0100)

----------------------------------------------------------------
pci, pc, virtio fixes and cleanups

Last large pull for soft freeze.

A bunch of fixes all over the place.
Most of ACPI refactoring has been merged.

virtio header cleanup has been merged,
now that windows has been build-tested with it.

initial patches from virtio-1.0 branch have been merged.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

----------------------------------------------------------------
Cornelia Huck (3):
      virtio: cull virtio_bus_set_vdev_features
      virtio: feature bit manipulation helpers
      virtio: add feature checking helpers

Igor Mammedov (43):
      acpi: move generic aml building helpers into dedictated file
      acpi: add build_append_namestring() helper
      acpi: drop min-bytes in build_package()
      pc: acpi-build: update linker on guest access
      pc: acpi-build: migrate RSDP table
      pc: acpi: use local var for accessing ACPI tables blob in acpi_build()
      acpi: introduce AML composer aml_append()
      acpi: add aml_scope() term
      pc: acpi-build: use aml_scope() for \_SB scope
      acpi: add aml_device() term
      acpi: add aml_method() term
      acpi: add aml_if() term
      acpi: add aml_name() & aml_name_decl() term
      acpi: add aml_int() term
      acpi: add aml_return() term
      acpi: add aml_arg() term
      acpi: add aml_store() term
      acpi: add aml_and() term
      acpi: add aml_notify() term
      acpi: add aml_call1(), aml_call2(), aml_call3(), aml_call4() helpers
      acpi: add aml_package() term
      pc: acpi-build: generate _S[345] packages dynamically
      acpi: add aml_buffer() term
      acpi: add aml_resource_template() helper
      acpi: add aml_io() helper
      acpi: include PkgLength size only when requested
      acpi: add aml_operation_region() term
      acpi: add aml_field() & aml_named_field() terms
      acpi: add aml_local() term
      acpi: add aml_string() term
      pc: acpi-build: generate pvpanic device description dynamically
      acpi: add aml_varpackage() term
      acpi: add aml_equal() term
      acpi: add aml_processor() term
      acpi: add aml_eisaid() term
      pc: acpi-build: drop template patching and CPU hotplug objects dynamically
      pc: acpi-build: create CPU hotplug IO region dynamically
      acpi: add aml_reserved_field() term
      pc: acpi-build: drop template patching and memory hotplug objects dynamically
      pc: acpi-build: create memory hotplug IO region dynamically
      acpi: add aml_word_bus_number(), aml_word_io(), aml_dword_memory(), aml_qword_memory() terms
      pc: pcihp: expose MMIO base and len as properties
      pc: acpi-build: reserve PCIHP MMIO resources

Markus Armbruster (11):
      pci: Convert core to realize
      pci: Permit incremental conversion of device models to realize
      pci: Trivial device model conversions to realize
      pcnet: pcnet_common_init() always returns 0, change to void
      pcnet: Convert to realize
      serial-pci: Convert to realize
      ide/ich: Convert to realize
      cirrus-vga: Convert to realize
      qxl: Convert to realize
      pci-assign: Convert to realize
      qdev: Don't exit when running into bad -global

Michael S. Tsirkin (32):
      acpi-build: fix memory leak with bridge hp off
      bios linker: validate pointer within table
      acpi: update RSDP on guest access
      exec: round up size on MR resize
      acpi-build: fix ACPI RAM management
      acpi: has_immutable_rsdp->!rsdp_in_ram
      acpi-build: simplify rsdp management for legacy
      scripts/update-linux-headers.sh: pull virtio hdrs
      include: import virtio headers from linux 4.0
      virtio: use standard virtio_ring.h
      virtio: use standard-headers
      virtio-balloon: use standard headers
      virtio-9p: use standard headers
      virtio-blk: switch to standard-headers
      virtio-net,tap: use standard-headers
      virtio-rng: use standard-headers
      virtio-scsi: use standard-headers
      virtio-serial: switch to standard-headers
      update-linux-headers: use standard-headers
      linux-headers: use standard-headers
      virtio-pci: use standard headers
      scripts: add arch specific standard-headers
      standard-headers: add s390 virtio headers
      s390: use standard headers
      acpi-build: skip hotplugged bridges
      update-linux-headers.sh: s/__inline__/inline/
      virtio-serial-bus.c: drop virtio_ids.h
      standard-headers: include stdint.h
      virtio_ring.h: s/__inline__/inline/
      acpi: update generated hex files
      acpi: drop unused generated files
      acpi-test: update expected files

Peter Krempa (2):
      vl.c: Fix error messages when parsing maxmem parameters
      pc: memory: Validate alignment of maxram_size to page size

Tang Chen (5):
      acpi, pc: Add hotunplug request cb for pc machine.
      acpi, ich9: Add hotunplug request cb for ich9.
      acpi, pc: Add unplug cb for pc machine.
      acpi, ich9: Add unplug cb for ich9.
      acpi, piix4: Add unplug cb for piix4.

 hw/9pfs/virtio-9p.h                                |  18 +-
 hw/net/pcnet.h                                     |   2 +-
 hw/s390x/s390-virtio-bus.h                         |  36 +-
 hw/s390x/s390-virtio.h                             |   7 +-
 include/hw/acpi/aml-build.h                        | 206 +++++
 include/hw/acpi/ich9.h                             |   4 +
 include/hw/acpi/pc-hotplug.h                       |   1 +
 include/hw/acpi/pcihp.h                            |   7 +-
 include/hw/i386/pc.h                               |   1 +
 include/hw/pci/pci.h                               |   3 +-
 include/hw/qdev-properties.h                       |   4 +-
 include/hw/virtio/dataplane/vring-accessors.h      |   2 +-
 include/hw/virtio/dataplane/vring.h                |   2 +-
 include/hw/virtio/virtio-balloon.h                 |  35 +-
 include/hw/virtio/virtio-blk.h                     |  77 +-
 include/hw/virtio/virtio-bus.h                     |   3 -
 include/hw/virtio/virtio-net.h                     | 151 +---
 include/hw/virtio/virtio-rng.h                     |   4 +-
 include/hw/virtio/virtio-scsi.h                    | 120 +--
 include/hw/virtio/virtio-serial.h                  |  40 +-
 include/hw/virtio/virtio.h                         |  71 +-
 include/net/tap.h                                  |  24 +-
 include/standard-headers/asm-s390/kvm_virtio.h     |  64 ++
 include/standard-headers/asm-s390/virtio-ccw.h     |  21 +
 include/standard-headers/linux/if_ether.h          |   1 +
 include/standard-headers/linux/types.h             |   2 +
 include/standard-headers/linux/virtio_9p.h         |  44 ++
 include/standard-headers/linux/virtio_balloon.h    |  59 ++
 include/standard-headers/linux/virtio_blk.h        | 143 ++++
 include/standard-headers/linux/virtio_config.h     |  64 ++
 include/standard-headers/linux/virtio_console.h    |  78 ++
 include/standard-headers/linux/virtio_ids.h        |  43 +
 include/standard-headers/linux/virtio_net.h        | 233 ++++++
 include/standard-headers/linux/virtio_pci.h        | 193 +++++
 .../linux}/virtio_ring.h                           | 132 ++--
 include/standard-headers/linux/virtio_rng.h        |   8 +
 include/standard-headers/linux/virtio_scsi.h       | 164 ++++
 include/standard-headers/linux/virtio_types.h      |  46 ++
 linux-headers/linux/virtio_config.h                |  58 +-
 linux-headers/linux/virtio_ring.h                  | 164 +---
 exec.c                                             |   2 +
 hw/9pfs/virtio-9p-device.c                         |   2 +-
 hw/acpi/aml-build.c                                | 867 +++++++++++++++++++++
 hw/acpi/bios-linker-loader.c                       |   4 +-
 hw/acpi/ich9.c                                     |  14 +
 hw/acpi/pcihp.c                                    |  18 +-
 hw/acpi/piix4.c                                    |  15 +-
 hw/audio/ac97.c                                    |   5 +-
 hw/audio/es1370.c                                  |   5 +-
 hw/audio/intel-hda.c                               |   6 +-
 hw/block/virtio-blk.c                              |  31 +-
 hw/char/serial-pci.c                               |  22 +-
 hw/char/virtio-serial-bus.c                        |   4 +-
 hw/core/qdev-properties.c                          |  21 +-
 hw/core/qdev.c                                     |   8 +-
 hw/display/cirrus_vga.c                            |  11 +-
 hw/display/qxl.c                                   |  36 +-
 hw/display/vga-pci.c                               |  11 +-
 hw/display/vmware_vga.c                            |   6 +-
 hw/i2c/smbus_ich9.c                                |   5 +-
 hw/i386/acpi-build.c                               | 750 +++++++++---------
 hw/i386/kvm/pci-assign.c                           |  10 +-
 hw/i386/pc.c                                       |  23 +
 hw/i386/pc_piix.c                                  |   3 +
 hw/i386/pc_q35.c                                   |   3 +
 hw/ide/cmd646.c                                    |   5 +-
 hw/ide/ich.c                                       |  13 +-
 hw/ide/piix.c                                      |  10 +-
 hw/ide/via.c                                       |   6 +-
 hw/ipack/tpci200.c                                 |   6 +-
 hw/isa/i82378.c                                    |   6 +-
 hw/isa/lpc_ich9.c                                  |  14 +-
 hw/isa/piix4.c                                     |   5 +-
 hw/isa/vt82c686.c                                  |  24 +-
 hw/misc/pci-testdev.c                              |   6 +-
 hw/net/e1000.c                                     |   6 +-
 hw/net/eepro100.c                                  |   6 +-
 hw/net/lance.c                                     |   3 +-
 hw/net/ne2000.c                                    |   6 +-
 hw/net/pcnet-pci.c                                 |   6 +-
 hw/net/pcnet.c                                     |   4 +-
 hw/net/rtl8139.c                                   |   6 +-
 hw/net/vhost_net.c                                 |   2 +-
 hw/net/virtio-net.c                                |  57 +-
 hw/net/vmxnet3.c                                   |   6 +-
 hw/pci-bridge/dec.c                                |   5 +-
 hw/pci-host/apb.c                                  |   5 +-
 hw/pci-host/bonito.c                               |   6 +-
 hw/pci-host/grackle.c                              |   5 +-
 hw/pci-host/piix.c                                 |  12 +-
 hw/pci-host/ppce500.c                              |   6 +-
 hw/pci-host/prep.c                                 |   6 +-
 hw/pci-host/q35.c                                  |   5 +-
 hw/pci-host/uninorth.c                             |  20 +-
 hw/pci-host/versatile.c                            |   5 +-
 hw/pci/pci.c                                       | 113 +--
 hw/s390x/virtio-ccw.c                              |   7 +-
 hw/scsi/virtio-scsi.c                              |   9 +-
 hw/sd/sdhci.c                                      |   5 +-
 hw/usb/hcd-ehci-pci.c                              |   6 +-
 hw/usb/hcd-xhci.c                                  |   6 +-
 hw/virtio/dataplane/vring.c                        |  10 +-
 hw/virtio/virtio-balloon.c                         |   2 +-
 hw/virtio/virtio-bus.c                             |  14 -
 hw/virtio/virtio-mmio.c                            |   2 +-
 hw/virtio/virtio-pci.c                             |  58 +-
 hw/virtio/virtio.c                                 |  33 +-
 hw/watchdog/wdt_i6300esb.c                         |   6 +-
 vl.c                                               |  34 +-
 hw/acpi/Makefile.objs                              |   1 +
 hw/i386/Makefile.objs                              |   6 +-
 hw/i386/acpi-dsdt-cpu-hotplug.dsl                  |  17 +-
 hw/i386/acpi-dsdt-mem-hotplug.dsl                  |  36 +-
 hw/i386/acpi-dsdt.hex.generated                    | 224 +-----
 hw/i386/q35-acpi-dsdt.hex.generated                | 224 +-----
 hw/i386/ssdt-mem.dsl                               |  77 --
 hw/i386/ssdt-mem.hex.generated                     | 213 -----
 hw/i386/ssdt-misc.dsl                              |  82 --
 hw/i386/ssdt-misc.hex.generated                    | 274 +------
 hw/i386/ssdt-proc.dsl                              |  63 --
 hw/i386/ssdt-proc.hex.generated                    | 134 ----
 scripts/update-linux-headers.sh                    |  52 +-
 tests/acpi-test-data/pc/DSDT                       | Bin 3621 -> 3415 bytes
 tests/acpi-test-data/pc/SSDT                       | Bin 2279 -> 2374 bytes
 tests/acpi-test-data/q35/DSDT                      | Bin 8211 -> 8005 bytes
 tests/acpi-test-data/q35/SSDT                      | Bin 560 -> 578 bytes
 126 files changed, 3237 insertions(+), 2960 deletions(-)
 create mode 100644 include/hw/acpi/aml-build.h
 create mode 100644 include/standard-headers/asm-s390/kvm_virtio.h
 create mode 100644 include/standard-headers/asm-s390/virtio-ccw.h
 create mode 100644 include/standard-headers/linux/if_ether.h
 create mode 100644 include/standard-headers/linux/types.h
 create mode 100644 include/standard-headers/linux/virtio_9p.h
 create mode 100644 include/standard-headers/linux/virtio_balloon.h
 create mode 100644 include/standard-headers/linux/virtio_blk.h
 create mode 100644 include/standard-headers/linux/virtio_config.h
 create mode 100644 include/standard-headers/linux/virtio_console.h
 create mode 100644 include/standard-headers/linux/virtio_ids.h
 create mode 100644 include/standard-headers/linux/virtio_net.h
 create mode 100644 include/standard-headers/linux/virtio_pci.h
 rename include/{hw/virtio => standard-headers/linux}/virtio_ring.h (61%)
 create mode 100644 include/standard-headers/linux/virtio_rng.h
 create mode 100644 include/standard-headers/linux/virtio_scsi.h
 create mode 100644 include/standard-headers/linux/virtio_types.h
 create mode 100644 hw/acpi/aml-build.c
 delete mode 100644 hw/i386/ssdt-mem.dsl
 delete mode 100644 hw/i386/ssdt-mem.hex.generated
 delete mode 100644 hw/i386/ssdt-proc.dsl
 delete mode 100644 hw/i386/ssdt-proc.hex.generated

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Peter Maydell]

On 27 February 2015 at 01:59, Michael S. Tsirkin <mst@redhat.com> wrote:
> Exactly the same thing as v1 but rebased to master to
> help Peter resolve conflicts, so I'm not bothering
> with resending all patches.
>
> A huge patchset, but the scariest part is Igor's patches,
> and these have been used by multiple people by now.
> virtio header change is a bit rushed, but getting them upstream
> seems like the best way to give them the cross-platform
> testing that they need, and this takes us a step closer to
> virtio-1.0 support.
>
> The following changes since commit 11d39a131020cc5c54ff9bc86d3259f7d32bf849:
>
>   Merge remote-tracking branch 'remotes/borntraeger/tags/s390x-20150218' into staging (2015-02-26 09:08:54 +0000)
>
> are available in the git repository at:
>
>   git://git.kernel.org/pub/scm/virt/kvm/mst/qemu.git tags/for_upstream_rebased
>
> for you to fetch changes up to dc33c0293505db9c747a6e27487a808ab2b62a6f:
>
>   acpi-test: update expected files (2015-02-26 13:04:19 +0100)
>
> ----------------------------------------------------------------
> pci, pc, virtio fixes and cleanups
>
> Last large pull for soft freeze.
>
> A bunch of fixes all over the place.
> Most of ACPI refactoring has been merged.
>
> virtio header cleanup has been merged,
> now that windows has been build-tested with it.
>
> initial patches from virtio-1.0 branch have been merged.
>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
>
> ----------------------------------------------------------------

Doesn't build with our minimum glib requirement, I'm afraid:

/home/petmay01/linaro/qemu-for-merges/hw/acpi/aml-build.c:313:
undefined reference to `g_ptr_array_new_with_free_func'

(that function only appeared in glib 2.22.)

-- PMM

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Paolo Bonzini]

On 03/03/2015 15:37, Peter Maydell wrote:
> Doesn't build with our minimum glib requirement, I'm afraid:
> 
> /home/petmay01/linaro/qemu-for-merges/hw/acpi/aml-build.c:313:
> undefined reference to `g_ptr_array_new_with_free_func'
> 
> (that function only appeared in glib 2.22.)

>From a few weeks ago:

> Moving beyond 2.22 would be awkward for me in that my OSX
> box only has 2.22 because fink doesn't have anything newer.
> I could probably deal with that somehow (switching to some
> other package system, probably).
> 
> Debian stable is "2.33.12+really2.32.4-5" and oldstable
> is "2.24.2-1" (and if my googling is right is an LTS release).
> 
> Ubuntu Lucid (LTS release) is 2.24; Precise (also LTS)
> is 2.32.
> 
> Daniel says RHEL6 has 2.28.
> 
> That suggests to me that we could reasonably advance to
> 2.22 or 2.24 if it seemed beneficial, but not beyond that.
> Is there anything particularly worthwhile that would get us?

Maybe we have a candidate now.  That function doesn't seem easy to
replace, and it is pretty useful.

Paolo

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Igor Mammedov]

On Tue, 03 Mar 2015 15:44:29 +0100
Paolo Bonzini <pbonzini@redhat.com> wrote:

> 
> 
> On 03/03/2015 15:37, Peter Maydell wrote:
> > Doesn't build with our minimum glib requirement, I'm afraid:
> > 
> > /home/petmay01/linaro/qemu-for-merges/hw/acpi/aml-build.c:313:
> > undefined reference to `g_ptr_array_new_with_free_func'
> > 
> > (that function only appeared in glib 2.22.)
> 
> From a few weeks ago:
> 
> > Moving beyond 2.22 would be awkward for me in that my OSX
> > box only has 2.22 because fink doesn't have anything newer.
> > I could probably deal with that somehow (switching to some
> > other package system, probably).
> > 
> > Debian stable is "2.33.12+really2.32.4-5" and oldstable
> > is "2.24.2-1" (and if my googling is right is an LTS release).
> > 
> > Ubuntu Lucid (LTS release) is 2.24; Precise (also LTS)
> > is 2.32.
> > 
> > Daniel says RHEL6 has 2.28.
> > 
> > That suggests to me that we could reasonably advance to
> > 2.22 or 2.24 if it seemed beneficial, but not beyond that.
> > Is there anything particularly worthwhile that would get us?
> 
> Maybe we have a candidate now.  That function doesn't seem easy to
> replace, and it is pretty useful.
Looking at git history there were a number of followup fixes
where 2.13-2.22 function couldn't be used, if we switch to 2.24
we could cleanup that places also.

> 
> Paolo
> 

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Paolo Bonzini]

On 03/03/2015 16:03, Igor Mammedov wrote:
>> > Maybe we have a candidate now.  That function doesn't seem easy to
>> > replace, and it is pretty useful.
> Looking at git history there were a number of followup fixes
> where 2.13-2.22 function couldn't be used, if we switch to 2.24
> we could cleanup that places also.

Peter mentioned above that 2.22 is a bit awkward to him, and he's
already swamped enough that we should consider doing him a favor. :)

Paolo

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Peter Maydell]

On 4 March 2015 at 00:05, Paolo Bonzini <pbonzini@redhat.com> wrote:
> On 03/03/2015 16:03, Igor Mammedov wrote:
>>> > Maybe we have a candidate now.  That function doesn't seem easy to
>>> > replace, and it is pretty useful.
>> Looking at git history there were a number of followup fixes
>> where 2.13-2.22 function couldn't be used, if we switch to 2.24
>> we could cleanup that places also.
>
> Peter mentioned above that 2.22 is a bit awkward to him, and he's
> already swamped enough that we should consider doing him a favor. :)

More to the point, Cornelia pointed out that SLES 11 SP 3 is still
on 2.22.

I guess I don't object to moving forward to 2.22 if somebody wants
to write the configure patch and add the info to the release notes.

-- PMM

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Paolo Bonzini]

On 03/03/2015 16:11, Peter Maydell wrote:
> I guess I don't object to moving forward to 2.22 if somebody wants
> to write the configure patch and add the info to the release notes.

Sent a patch for the former.

Paolo

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Igor Mammedov]

On Wed, 4 Mar 2015 00:11:56 +0900
Peter Maydell <peter.maydell@linaro.org> wrote:

> On 4 March 2015 at 00:05, Paolo Bonzini <pbonzini@redhat.com> wrote:
> > On 03/03/2015 16:03, Igor Mammedov wrote:
> >>> > Maybe we have a candidate now.  That function doesn't seem easy to
> >>> > replace, and it is pretty useful.
> >> Looking at git history there were a number of followup fixes
> >> where 2.13-2.22 function couldn't be used, if we switch to 2.24
> >> we could cleanup that places also.
> >
> > Peter mentioned above that 2.22 is a bit awkward to him, and he's
> > already swamped enough that we should consider doing him a favor. :)
> 
> More to the point, Cornelia pointed out that SLES 11 SP 3 is still
> on 2.22.
> 
> I guess I don't object to moving forward to 2.22 if somebody wants
> to write the configure patch and add the info to the release notes.
I've just sent patch on list as for release notes is it here:
http://wiki.qemu.org/ChangeLog/2.3

> 
> -- PMM

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Alex Bennée]

Peter Maydell <peter.maydell@linaro.org> writes:

> On 4 March 2015 at 00:05, Paolo Bonzini <pbonzini@redhat.com> wrote:
>> On 03/03/2015 16:03, Igor Mammedov wrote:
>>>> > Maybe we have a candidate now.  That function doesn't seem easy to
>>>> > replace, and it is pretty useful.
>>> Looking at git history there were a number of followup fixes
>>> where 2.13-2.22 function couldn't be used, if we switch to 2.24
>>> we could cleanup that places also.
>>
>> Peter mentioned above that 2.22 is a bit awkward to him, and he's
>> already swamped enough that we should consider doing him a favor. :)
>
> More to the point, Cornelia pointed out that SLES 11 SP 3 is still
> on 2.22.
>
> I guess I don't object to moving forward to 2.22 if somebody wants
> to write the configure patch and add the info to the release notes.

The commit comment for configure:

    glib_req_ver=2.12

States it's the version found in RHEL5. Do we not care about that
anymore? I guess at some point the enterprise distros are only
backporting fixes to an old qemu anyway.

(/me is all for newer glibs btw, there are some handy features)
>
> -- PMM

-- 
Alex Bennée

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Paolo Bonzini]

On 03/03/2015 16:55, Alex Bennée wrote:
> The commit comment for configure:
> 
>     glib_req_ver=2.12
> 
> States it's the version found in RHEL5. Do we not care about that
> anymore? I guess at some point the enterprise distros are only
> backporting fixes to an old qemu anyway.

This was not chosen for the sake of Red Hat.  It's for people who need
to do emulation but are stuck with RHEL5 because of some EDA tool they use.

Paolo

> (/me is all for newer glibs btw, there are some handy features)

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Michael S. Tsirkin]

On Tue, Mar 03, 2015 at 03:44:29PM +0100, Paolo Bonzini wrote:
> 
> 
> On 03/03/2015 15:37, Peter Maydell wrote:
> > Doesn't build with our minimum glib requirement, I'm afraid:
> > 
> > /home/petmay01/linaro/qemu-for-merges/hw/acpi/aml-build.c:313:
> > undefined reference to `g_ptr_array_new_with_free_func'
> > 
> > (that function only appeared in glib 2.22.)
> 
> >From a few weeks ago:
> 
> > Moving beyond 2.22 would be awkward for me in that my OSX
> > box only has 2.22 because fink doesn't have anything newer.
> > I could probably deal with that somehow (switching to some
> > other package system, probably).
> > 
> > Debian stable is "2.33.12+really2.32.4-5" and oldstable
> > is "2.24.2-1" (and if my googling is right is an LTS release).
> > 
> > Ubuntu Lucid (LTS release) is 2.24; Precise (also LTS)
> > is 2.32.
> > 
> > Daniel says RHEL6 has 2.28.
> > 
> > That suggests to me that we could reasonably advance to
> > 2.22 or 2.24 if it seemed beneficial, but not beyond that.
> > Is there anything particularly worthwhile that would get us?
> 
> Maybe we have a candidate now.  That function doesn't seem easy to
> replace, and it is pretty useful.
> 
> Paolo


It's my mistake.
It does seem trivial to replace:

-->

aml-build: fix build for glib < 2.22

g_ptr_array_new_with_free_func is there since glib 2.22,
use the older g_ptr_array_foreach instead.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

---

diff --git a/hw/acpi/aml-build.c b/hw/acpi/aml-build.c
index 3e5949b..f46c96c 100644
--- a/hw/acpi/aml-build.c
+++ b/hw/acpi/aml-build.c
@@ -299,7 +299,7 @@ static Aml *aml_bundle(uint8_t op, AmlBlockFlags flags)
     return var;
 }
 
-static void aml_free(gpointer data)
+static void aml_free(gpointer data, gpointer user_data)
 {
     Aml *var = data;
     build_free_array(var->buf);
@@ -310,13 +310,14 @@ Aml *init_aml_allocator(void)
     Aml *var;
 
     assert(!alloc_list);
-    alloc_list = g_ptr_array_new_with_free_func(aml_free);
+    alloc_list = g_ptr_array_new();
     var = aml_alloc();
     return var;
 }
 
 void free_aml_allocator(void)
 {
+    g_ptr_array_foreach(alloc_list, aml_free, NULL)
     g_ptr_array_free(alloc_list, true);
     alloc_list = 0;
 }

Re: [Qemu-devel] [PULL v2] pci, pc, virtio fixes and cleanups

[Michael S. Tsirkin]

On Tue, Mar 03, 2015 at 11:37:27PM +0900, Peter Maydell wrote:
> On 27 February 2015 at 01:59, Michael S. Tsirkin <mst@redhat.com> wrote:
> > Exactly the same thing as v1 but rebased to master to
> > help Peter resolve conflicts, so I'm not bothering
> > with resending all patches.
> >
> > A huge patchset, but the scariest part is Igor's patches,
> > and these have been used by multiple people by now.
> > virtio header change is a bit rushed, but getting them upstream
> > seems like the best way to give them the cross-platform
> > testing that they need, and this takes us a step closer to
> > virtio-1.0 support.
> >
> > The following changes since commit 11d39a131020cc5c54ff9bc86d3259f7d32bf849:
> >
> >   Merge remote-tracking branch 'remotes/borntraeger/tags/s390x-20150218' into staging (2015-02-26 09:08:54 +0000)
> >
> > are available in the git repository at:
> >
> >   git://git.kernel.org/pub/scm/virt/kvm/mst/qemu.git tags/for_upstream_rebased
> >
> > for you to fetch changes up to dc33c0293505db9c747a6e27487a808ab2b62a6f:
> >
> >   acpi-test: update expected files (2015-02-26 13:04:19 +0100)
> >
> > ----------------------------------------------------------------
> > pci, pc, virtio fixes and cleanups
> >
> > Last large pull for soft freeze.
> >
> > A bunch of fixes all over the place.
> > Most of ACPI refactoring has been merged.
> >
> > virtio header cleanup has been merged,
> > now that windows has been build-tested with it.
> >
> > initial patches from virtio-1.0 branch have been merged.
> >
> > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> >
> > ----------------------------------------------------------------
> 
> Doesn't build with our minimum glib requirement, I'm afraid:
> 
> /home/petmay01/linaro/qemu-for-merges/hw/acpi/aml-build.c:313:
> undefined reference to `g_ptr_array_new_with_free_func'
> 
> (that function only appeared in glib 2.22.)
> 
> -- PMM

I fixed that, PULL v4 is out.
Sorry about all the noise.

-- 
MST