qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v8 00/29] Consolidate PIIX south bridges


From: Bernhard Beschow
Subject: Re: [PATCH v8 00/29] Consolidate PIIX south bridges
Date: Wed, 11 Oct 2023 18:57:07 +0000


Am 8. Oktober 2023 17:56:48 UTC schrieb Chuck Zmudzinski <brchuckz@aol.com>:
>On 10/7/23 8:38 AM, Bernhard Beschow wrote:
>> This series consolidates the implementations of the PIIX3 and PIIX4 south
>> bridges and makes PIIX4 usable in the PC machine via an experimental command
>> line parameter. The motivation is to resolve duplicate code between the 
>> device
>> models as well as resolving the "Frankenstein" PIIX4-PM problem in PIIX3
>> discussed on this list before.
>> 
>> The series is structured as follows:
>> 
>> Patches 1-8 are preparational patches necessary for moving all sub devices 
>> into
>> PIIX3, like was done for PIIX4. In isolation these patches can also be seen 
>> as
>> general x86 machine cleanup sub series which has merit in its own right -- 
>> and
>> could be applied to master if the remainder of the series takes longer to
>> review.
>> 
>> Patches 9-13 move PIIX3 sub devices into one device model like already
>> done for PIIX4. Together with the previous sub series these patches form a
>> bigger sub series which also has merit in its own right, and could be applied
>> independent of the remainder of this series as well.
>> 
>> The remainder of this series consolidates the PIIX3 and PIIX4 device models.
>> The culmination point is the last commit which makes PIIX4 usable in the PC
>> machine.
>> 
>> One challenge was dealing with optional devices where Peter already gave 
>> advice
>> in [1] which this series implements. Although PIIX4 is now usable in the PC
>> machine it still has a different binary layout in its VM state.
>> 
>> Testing done:
>> * `make check`
>> * `qemu-system-x86_64 -M pc -m 2G -accel kvm -cdrom
>>      manjaro-kde-21.3.2-220704-linux515.iso`
>> * `qemu-system-x86_64 -M pc,x-south-bridge=piix4-isa -m 2G -accel kvm -cdrom
>>      manjaro-kde-21.3.2-220704-linux515.iso`
>> * `qemu-system-x86_64 -M q35 -m 2G -accel kvm -cdrom
>>      manjaro-kde-21.3.2-220704-linux515.iso`
>> * `qemu-system-mips64el -M malta -cpu 5KEc -m 1G -kernel kernel -initrd 
>> initrd
>>      -append "root=LABEL=rootfs console=ttyS0" -drive file=image.qcow2`
>> * `qemu-system-mips64el -M malta -bios yamon-02.22.bin`
>> * Run HVM domU guest under Xen with manjaro-kde-21.3.2-220704-linux515.iso 
>> image
>
>I did some preliminary tests of this patch series on some Xen HVM domU guests 
>I have
>that use the xenfv / pc machine and depend on the current PIIX3 implementation.
>So far there are no regressions in my tests. I use libxl or libvirt to manage 
>the
>Xen guests.

Thanks, nice to read!

>
>I have not (yet) tested the experimental option that makes PIIX4 useable in 
>the xenfv / pc
>machines. IIUC, that would require a patch to hvmloader/pci.c in Xen tools so 
>Xen's
>hvmloader recognizes the PIIX4 pci device id [1], and a patch to libxl so 
>libxl can
>optionally launch qemu with the new experimental option enabled.
>
>Since this patch series affects the xenfv machine, I added the Xen x86 
>maintainers to
>the Cc list and Jason Andryuk who is credited with discovering the necessary 
>patch to
>hvmloader/pci.c.

Good idea. In the next iteration, I'll cc the respective email addresses from 
the MAINTAINERS file which hopefully reaches all relevant people.

Best regards,
Bernhard

>
>[1] 
>https://lore.kernel.org/qemu-devel/B0FF78F4-1193-495B-919C-84A1FF8ADF12@gmail.com/
>
>> 
>> v8:
>> - Wire ISA interrupts before device realization
>> - Optionally allow a PIC and PIT to be instantiated in PIIX3 for compatiblity
>>     with PIIX4
>> - Touch ICH9 LPC as far as required for PIIX consolidation
>> - Make PIIX4 usable in the PC machine via an experimental option
>> - Review and rework history, touching every commit and drop R-b tags when
>>     changes became too large
>> 
>> v7:
>> - Rebase onto master
>> - Avoid the PIC proxy (Phil)
>>   The motivation for the PIC proxy was to allow for wiring up ISA interrupts 
>> in
>>   the south bridges. ISA interrupt wiring requires the GPIO lines to be
>>   populated already but pc_piix assigned the interrupts only after realizing
>>   PIIX3. By shifting interrupt assignment before realizing, the ISA 
>> interrupts
>>   are already populated during PIIX3's realize phase where the ISA interrupts
>>   are wired up.
>> - New patches:
>>   * hw/isa/piix4: Reuse struct PIIXState from PIIX3
>>   * hw/isa/piix4: Create the "intr" property during init() already
>> - Patches with substantial changes (Reviewed-by dropped):
>>   * hw/isa/piix3: Move ISA bus IRQ assignments into host device
>> 
>> v6:
>> - Fix some comments about TYPE_ISA_PIC (Mark) ... and use it consistently
>>   within the patch series.
>> - Incorporate series "[PATCH v2 0/3] Decouple INTx-to-LNKx routing from south
>>   bridges" [2] for maintainer convenience.
>> - Merge v5's 'hw/i386/pc_piix: Associate pci_map_irq_fn as soon as PCI bus is
>>   created' into
>>   https://lists.nongnu.org/archive/html/qemu-devel/2022-11/msg03312.html . Do
>>   similar for Malta.
>> - Rebase onto latest master (d6271b657286 "Merge tag 'for_upstream' of
>>   https://git.kernel.org/pub/scm/virt/kvm/mst/qemu into staging")
>> 
>> v5:
>> - Pick up Reviewed-by tags from 
>> https://lists.nongnu.org/archive/html/qemu-devel/2023-01/msg00116.html
>> - Add patch to make usage of the isa_pic global more type-safe
>> - Re-introduce isa-pic as PIC specific proxy (Mark)
>> 
>> v4:
>> - Rebase onto "[PATCH v2 0/3] Decouple INTx-to-LNKx routing from south 
>> bridges"
>>   since it is already queued via mips-next. This eliminates patches
>>   'hw/isa/piix3: Prefix pci_slot_get_pirq() with "piix3_"' and 'hw/isa/piix4:
>>   Prefix pci_slot_get_pirq() with "piix4_"'.
>> - Squash 'hw/isa/piix: Drop the "3" from the PIIX base class' into
>>   'hw/isa/piix3: Rename typedef PIIX3State to PIIXState'. I originally only
>>   split these patches since I wasn't sure whether renaming a type was 
>> allowed.
>> - Add new patch 'hw/i386/pc_piix: Associate pci_map_irq_fn as soon as PCI 
>> bus is
>>   created' for forther cleanup of INTx-to-LNKx route decoupling.
>> 
>> v3:
>> - Introduce one TYPE_ICH9_USB_UHCI(fn) rather than several 
>> TYPE_ICH9_USB_UHCIx
>>   (Philippe)
>> - Make proxy PIC generic (Philippe)
>> - Track Malta's PIIX dependencies through KConfig
>> - Rebase onto Philippe's 'hw/isa/piix4: Remove MIPS Malta specific bits' 
>> series [3]
>> - Also rebase onto latest master to resolve merge conflicts. This required
>>   copying Philippe's series as first three patches - please ignore.
>> 
>> v2:
>> - Introduce TYPE_ defines for IDE and USB device models (Mark)
>> - Omit unexporting of PIIXState (Mark)
>> - Improve commit message of patch 5 to mention reset triggering through PCI
>>   configuration space (Mark)
>> - Move reviewed patches w/o dependencies to the bottom of the series for 
>> early
>>   upstreaming
>> 
>> [1] https://lists.nongnu.org/archive/html/qemu-devel/2022-07/msg02348.html
>> [2] https://lists.nongnu.org/archive/html/qemu-devel/2022-11/msg03310.html
>> [3] https://lists.nongnu.org/archive/html/qemu-devel/2022-10/msg05367.html
>> 
>> Bernhard Beschow (29):
>>   hw/i386/pc: Merge two if statements into one
>>   hw/i386/pc_piix: Allow for setting properties before realizing PIIX3
>>     south bridge
>>   hw/i386/pc_piix: Assign PIIX3's ISA interrupts before its realize()
>>   hw/isa/piix3: Resolve redundant PIIX_NUM_PIC_IRQS
>>   hw/i386/pc_piix: Wire PIIX3's ISA interrupts by new "isa-irqs"
>>     property
>>   hw/i386/pc_piix: Remove redundant "piix3" variable
>>   hw/isa/piix3: Rename "pic" attribute to "isa_irqs_in"
>>   hw/i386/pc_q35: Wire ICH9 LPC function's interrupts before its
>>     realize()
>>   hw/isa/piix3: Wire PIC IRQs to ISA bus in host device
>>   hw/i386/pc: Wire RTC ISA IRQs in south bridges
>>   hw/isa/piix3: Create IDE controller in host device
>>   hw/isa/piix3: Create USB controller in host device
>>   hw/isa/piix3: Create power management controller in host device
>>   hw/isa/piix3: Drop the "3" from PIIX base class name
>>   hw/isa/piix4: Remove unused inbound ISA interrupt lines
>>   hw/isa/piix4: Rename "isa" attribute to "isa_irqs_in"
>>   hw/isa/piix4: Rename reset control operations to match PIIX3
>>   hw/isa/piix4: Reuse struct PIIXState from PIIX3
>>   hw/isa/piix3: Merge hw/isa/piix4.c
>>   hw/isa/piix: Allow for optional PIC creation in PIIX3
>>   hw/isa/piix: Allow for optional PIT creation in PIIX3
>>   hw/isa/piix: Harmonize names of reset control memory regions
>>   hw/isa/piix: Share PIIX3's base class with PIIX4
>>   hw/isa/piix: Reuse PIIX3 base class' realize method in PIIX4
>>   hw/isa/piix: Rename functions to be shared for PCI interrupt
>>     triggering
>>   hw/isa/piix: Reuse PIIX3's PCI interrupt triggering in PIIX4
>>   hw/isa/piix: Resolve duplicate code regarding PCI interrupt wiring
>>   hw/isa/piix: Implement multi-process QEMU support also for PIIX4
>>   hw/i386/pc_piix: Make PIIX4 south bridge usable in PC machine
>> 
>>  MAINTAINERS                          |   6 +-
>>  docs/system/target-i386-desc.rst.inc |   8 +
>>  include/hw/i386/pc.h                 |   2 +
>>  include/hw/southbridge/piix.h        |  28 ++-
>>  hw/i386/pc.c                         |  13 +-
>>  hw/i386/pc_piix.c                    | 125 ++++++++---
>>  hw/i386/pc_q35.c                     |  14 +-
>>  hw/isa/lpc_ich9.c                    |   9 +-
>>  hw/isa/{piix3.c => piix.c}           | 281 ++++++++++++++++++-------
>>  hw/isa/piix4.c                       | 302 ---------------------------
>>  hw/mips/malta.c                      |   5 +-
>>  hw/i386/Kconfig                      |   3 +-
>>  hw/isa/Kconfig                       |   8 +-
>>  hw/isa/meson.build                   |   3 +-
>>  hw/mips/Kconfig                      |   2 +-
>>  15 files changed, 358 insertions(+), 451 deletions(-)
>>  rename hw/isa/{piix3.c => piix.c} (52%)
>>  delete mode 100644 hw/isa/piix4.c
>> 
>



reply via email to

[Prev in Thread] Current Thread [Next in Thread]