On Tue, Jan 20, 2015 at 09:38:18PM +0100, Igor Mammedov wrote: > On Tue, 20 Jan 2015 18:29:07 +0200 > "Michael S. Tsirkin" <m...@redhat.com> wrote: > > > On Tue, Jan 20, 2015 at 03:52:25PM +0000, Igor Mammedov wrote: > > > NOTE to maintainer: please update test data (ACPI blobs) in test > > > cases > > > > > > changes from v3: > > > * rename acpi_gen_utils.[ch] to acpi-build-utils.[ch] > > > * copy GLP license block from acpi-build.c > > > * assert on wrong Segcount earlier and extend condition to > > > seg_count > 0 && seg_count <= 255 > > > * drop "pc: acpi: decribe bridge device as not hotpluggable" > > > * keep original logic of creating bridge devices as it was done > > > in 133a2da48 "pc: acpi: generate AML only for PCI0 ..." > > > * if bus is non hotpluggable, add child slots to bus as non > > > hotpluggable as it was done in original code. > > > > > > changes from v2: > > > * codding style fixups > > > * check for SegCount earlier > > > * use hotpluggable device object instead of not hotpluggable > > > for non present devices, and add it only when bus itself is > > > hotpluggable > > > > > > changes from v1: > > > * drop: [PATCH 7/9] acpi: replace opencoded notify codes with > > > named values > > > * use Michael's suggestion to improve build_append_nameseg() > > > * drop long scope names and go back to recursion, > > > but still significantly simplify building of PCI tree > > > > > > this series is an attempt to shave off a bunch of > > > not directly related patches from already big dynamic > > > AML series (although it's dependency for it) > > > > > > Tested: on XPsp3 to WS2012R2 and REHL6/7 guests. > > > > > > Git tree for testing: > > > https://github.com/imammedo/qemu/commits/acpi_pci_gen_simplification_v4 > > > > > > previous revision: > > > [PATCH v3 0/8] pc: acpi: various fixes and cleanups > > > http://lists.gnu.org/archive/html/qemu-devel/2014-12/msg03057.html > > > > > > Igor Mammedov (7): > > > pc: acpi: fix WindowsXP BSOD when memory hotplug is enabled > > > > At least this one is already in pci tree isn't it? > > Could you pls rebase? > > Or if there's a replacement patch, pls send fixup or > > mention old one needs to be retracted in commit log. > I've rebased it on PCI tree, there was a trivial conflict in 4/7 > you can pull result from: > https://github.com/imammedo/qemu/commits/acpi_pci_gen_simplification_v4_PCI
That's not a git url. Please just repost, it's easier. > > > > > > > > > pc: acpi-build: cleanup AcpiPmInfo initialization > > > acpi: build_append_nameseg(): add padding if necessary > > > 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: simplify PCI bus tree generation > > > > > > hw/acpi/Makefile.objs | 1 + > > > hw/acpi/acpi-build-utils.c | 269 +++++++++++++++++++++ > > > hw/i386/acpi-build.c | 464 > > > +++++++++---------------------------- > > > hw/i386/acpi-dsdt-cpu-hotplug.dsl | 1 + > > > include/hw/acpi/acpi-build-utils.h | 23 ++ 5 files changed, 398 > > > insertions(+), 360 deletions(-) create mode 100644 > > > hw/acpi/acpi-build-utils.c create mode 100644 > > > include/hw/acpi/acpi-build-utils.h > > > > > > -- > > > 1.8.3.1
