]> git.ipfire.org Git - thirdparty/qemu.git/log
thirdparty/qemu.git
6 days agohw/arm: Integrate ASPEED OTP memory support into AST1030 SoCs
Kane-Chen-AS [Tue, 12 Aug 2025 09:40:03 +0000 (17:40 +0800)] 
hw/arm: Integrate ASPEED OTP memory support into AST1030 SoCs

The has_otp attribute is enabled in the SBC subclasses for AST1030 to
control the presence of OTP support per SoC type.

Signed-off-by: Kane-Chen-AS <kane_chen@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250812094011.2617526-7-kane_chen@aspeedtech.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
6 days agohw/nvram/aspeed_otp: Add OTP programming semantics and tracing
Kane-Chen-AS [Tue, 12 Aug 2025 09:40:02 +0000 (17:40 +0800)] 
hw/nvram/aspeed_otp: Add OTP programming semantics and tracing

Implement correct OTP programming behavior for Aspeed OTP:
- Support read-modify-write flow with one-way bit programming:
  * prog_bit uses 0s as the "to-be-programmed" mask.
  * Even-indexed words: 0->1, odd-indexed words: 1->0.
  * Reject non-programmable requests and log conflicts.
- Enable unaligned accesses in MemoryRegionOps.
  Since each OTP address maps to a 1DW (4B) or 2DW (8B) block in the
  backing store, upper-layer accesses may be unaligned to block
  boundaries.

This matches the irreversible, word-parity-dependent programming rules
of Aspeed SoCs and exposes changes via QEMU trace events.

Signed-off-by: Kane-Chen-AS <kane_chen@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250812094011.2617526-6-kane_chen@aspeedtech.com
[ clg: Fixed PRIx64 format in aspeed_otp_write() ]
Signed-off-by: Cédric Le Goater <clg@redhat.com>
6 days agohw/nvram/aspeed_otp: Add 'drive' property to support block backend
Kane-Chen-AS [Tue, 12 Aug 2025 09:40:01 +0000 (17:40 +0800)] 
hw/nvram/aspeed_otp: Add 'drive' property to support block backend

This patch introduces a 'drive' property to the Aspeed OTP device,
allowing it to be backed by a block device. Users can now preload
OTP data via QEMU CLI using a block backend.

Example usage:
  ./qemu-system-arm \
    -blockdev driver=file,filename=otpmem.img,node-name=otp \
    -global aspeed-otp.drive=otp \
    ...

If the drive is provided, its content will be loaded as the initial OTP
state. Otherwise, an internal memory buffer will be used.

Signed-off-by: Kane-Chen-AS <kane_chen@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250812094011.2617526-5-kane_chen@aspeedtech.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
6 days agohw/arm: Integrate ASPEED OTP memory support into AST2600 SoCs
Kane-Chen-AS [Tue, 12 Aug 2025 09:40:00 +0000 (17:40 +0800)] 
hw/arm: Integrate ASPEED OTP memory support into AST2600 SoCs

The has_otp attribute is enabled in the SBC subclasses for AST2600 to
control the presence of OTP support per SoC type.

Signed-off-by: Kane-Chen-AS <kane_chen@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250812094011.2617526-4-kane_chen@aspeedtech.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
6 days agohw/misc/aspeed_sbc: Connect ASPEED OTP memory device to SBC
Kane-Chen-AS [Tue, 12 Aug 2025 09:39:59 +0000 (17:39 +0800)] 
hw/misc/aspeed_sbc: Connect ASPEED OTP memory device to SBC

This patch connects the aspeed.otp device to the ASPEED Secure Boot
Controller (SBC) model. It implements OTP memory access via the SBC's
command interface and enables emulation of secure fuse programming
flows.

The following OTP commands are supported:
  - READ: reads a 32-bit word from OTP memory into internal registers
  - PROG: programs a 32-bit word value to the specified OTP address

Trace events are added to observe read/program operations and command
handling flow.

Signed-off-by: Kane-Chen-AS <kane_chen@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250812094011.2617526-3-kane_chen@aspeedtech.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
6 days agohw/nvram/aspeed_otp: Add ASPEED OTP memory device model
Kane-Chen-AS [Tue, 12 Aug 2025 09:39:58 +0000 (17:39 +0800)] 
hw/nvram/aspeed_otp: Add ASPEED OTP memory device model

Introduce a QEMU device model for ASPEED's One-Time Programmable (OTP)
memory.

This model simulates a word-addressable OTP region used for secure
fuse storage. The OTP memory can operate with an internal memory
buffer.

The OTP model provides a memory-like interface through a dedicated
AddressSpace, allowing other device models (e.g., SBC) to issue
transactions as if accessing a memory-mapped region.

Signed-off-by: Kane-Chen-AS <kane_chen@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250812094011.2617526-2-kane_chen@aspeedtech.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
7 days agoMerge tag 'pull-loongarch-20250928' of https://github.com/bibo-mao/qemu into staging
Richard Henderson [Sun, 28 Sep 2025 16:01:35 +0000 (09:01 -0700)] 
Merge tag 'pull-loongarch-20250928' of https://github.com/bibo-mao/qemu into staging

loongarch queue

# -----BEGIN PGP SIGNATURE-----
#
# iHUEABYKAB0WIQQNhkKjomWfgLCz0aQfewwSUazn0QUCaNjtuwAKCRAfewwSUazn
# 0Z9VAQDuqEzBEj0I3L7AtJgwRxSau+sw9FqUdAjQguM9mA29ggD7BOBFwHpjx68t
# 8MMstQuZN2mFRwzfukIdLDZclPCKkAM=
# =L9oL
# -----END PGP SIGNATURE-----
# gpg: Signature made Sun 28 Sep 2025 01:11:39 AM PDT
# gpg:                using EDDSA key 0D8642A3A2659F80B0B3D1A41F7B0C1251ACE7D1
# gpg: Good signature from "bibo mao <maobibo@loongson.cn>" [unknown]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 7044 3A00 19C0 E97A 31C7  13C4 8E86 8FB7 A176 9D4C
#      Subkey fingerprint: 0D86 42A3 A265 9F80 B0B3  D1A4 1F7B 0C12 51AC E7D1

* tag 'pull-loongarch-20250928' of https://github.com/bibo-mao/qemu:
  target/loongarch: Only flush one TLB entry in helper_invtlb_page_asid()
  target/loongarch: Only flush one TLB entry in helper_invtlb_page_asid_or_g()
  target/loongarch: Invalid tlb entry in invalidate_tlb()
  target/loongarch: Use loongarch_tlb_search_cb in helper_invtlb_page_asid
  target/loongarch: Use loongarch_tlb_search_cb in helper_invtlb_page_asid_or_g
  target/loongarch: Change return value type with loongarch_tlb_search_cb()
  target/loongarch: Add common API loongarch_tlb_search_cb()
  target/loongarch: Add tlb search callback in loongarch_tlb_search()
  target/loongarch: Fix page size set issue with CSR_STLBPS
  target/loongarch: Update TLB index selection method
  target/loongarch: Reduce TLB flush with helper_tlbwr
  target/loongarch: Add parameter tlb pointer with fill_tlb_entry
  target/loongarch: Use mmu idx bitmap method when flush TLB

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
7 days agoMerge tag 'pull-loongarch-20250928' of https://github.com/gaosong715/qemu into staging
Richard Henderson [Sun, 28 Sep 2025 16:00:36 +0000 (09:00 -0700)] 
Merge tag 'pull-loongarch-20250928' of https://github.com/gaosong715/qemu into staging

pull-loongarch-20250928

v2: fix build win64 errors.

# -----BEGIN PGP SIGNATURE-----
#
# iLMEAAEIAB0WIQTKRzxE1qCcGJoZP81FK5aFKyaCFgUCaNkDHQAKCRBFK5aFKyaC
# Fn06A/0SQKLVcktq2lX+aRurdGw/LKt/1mtSFJes6s5VVCrNuFFzmkXzjs/m0CcX
# scgDF67Z+PhJpLtNLRV8FiJ+z3bOH/j+yRHqj1xnvvITb+i5bUYbt+A81wrzX6Bi
# J/Ayqu49oQj33hX3lqTcTBmwYDBc2v7nu0PfvFqOUi9bTvYgfA==
# =C4NB
# -----END PGP SIGNATURE-----
# gpg: Signature made Sun 28 Sep 2025 02:42:53 AM PDT
# gpg:                using RSA key CA473C44D6A09C189A193FCD452B96852B268216
# gpg: Good signature from "Song Gao <gaosong@loongson.cn>" [unknown]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: CA47 3C44 D6A0 9C18 9A19  3FCD 452B 9685 2B26 8216

* tag 'pull-loongarch-20250928' of https://github.com/gaosong715/qemu:
  hw/loongarch: Implement DINTC plug/unplug interfaces
  target/loongarch:Implement csrrd CSR_MSGIR register
  target/loongarch: Add CSR_ESTAT.bit15 and CSR_ECFG.bit15 for msg interrupts.
  hw/loongarch: Implement dintc set irq
  hw/loongarch: Implement dintc realize and unrealize
  hw/loongarch: DINTC add a MemoryRegion
  target/loongarch: add msg interrupt CSR registers
  loongarch: add a direct interrupt controller device
  hw/loongarch: add misc register support dmsi
  hw/loongarch: add virt feature dmsi support
  target/loongarch: move some machine define to virt.h

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
8 days agohw/loongarch: Implement DINTC plug/unplug interfaces
Song Gao [Tue, 16 Sep 2025 12:21:09 +0000 (20:21 +0800)] 
hw/loongarch: Implement DINTC plug/unplug interfaces

when cpu added, connect dintc irq to cpu INT_DMSI irq pin.

Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
Message-ID: <20250916122109.749813-12-gaosong@loongson.cn>

8 days agotarget/loongarch:Implement csrrd CSR_MSGIR register
Song Gao [Tue, 16 Sep 2025 12:21:08 +0000 (20:21 +0800)] 
target/loongarch:Implement csrrd CSR_MSGIR register

implement the read-clear feature for CSR_MSGIR register.

Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
Message-ID: <20250916122109.749813-11-gaosong@loongson.cn>

8 days agotarget/loongarch: Add CSR_ESTAT.bit15 and CSR_ECFG.bit15 for msg interrupts.
Song Gao [Tue, 16 Sep 2025 12:21:07 +0000 (20:21 +0800)] 
target/loongarch: Add CSR_ESTAT.bit15 and CSR_ECFG.bit15 for msg interrupts.

Add CSR_ESTAT.bit15 and CSR_ECFG.bit15 for DINTC irq.

Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
Message-ID: <20250916122109.749813-10-gaosong@loongson.cn>

8 days agohw/loongarch: Implement dintc set irq
Song Gao [Tue, 16 Sep 2025 12:21:06 +0000 (20:21 +0800)] 
hw/loongarch: Implement dintc set irq

Implement dintc set irq and update CSR_MSGIS.

Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
Message-ID: <20250916122109.749813-9-gaosong@loongson.cn>

8 days agohw/loongarch: Implement dintc realize and unrealize
Song Gao [Tue, 16 Sep 2025 12:21:05 +0000 (20:21 +0800)] 
hw/loongarch: Implement dintc realize and unrealize

Implement th DINTC realize and unrealize.

Signed-off-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Message-ID: <20250916122109.749813-8-gaosong@loongson.cn>

8 days agohw/loongarch: DINTC add a MemoryRegion
Song Gao [Tue, 16 Sep 2025 12:21:04 +0000 (20:21 +0800)] 
hw/loongarch: DINTC add a MemoryRegion

the DINTC use [2fe00000-2ff00000) Memory.

Signed-off-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Message-ID: <20250916122109.749813-7-gaosong@loongson.cn>

8 days agotarget/loongarch: add msg interrupt CSR registers
Song Gao [Tue, 16 Sep 2025 12:21:03 +0000 (20:21 +0800)] 
target/loongarch: add msg interrupt CSR registers

include CSR_MSGIS0-3, CSR_MSGIR and CSR_MSGIE.

Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Signed-off-by: Song Gao <gaosong@loongson.cn>
Message-ID: <20250916122109.749813-6-gaosong@loongson.cn>

8 days agoloongarch: add a direct interrupt controller device
Song Gao [Tue, 16 Sep 2025 12:21:02 +0000 (20:21 +0800)] 
loongarch: add a direct interrupt controller device

Add Loongarch direct interrupt controller device base Definition.

Signed-off-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Message-ID: <20250916122109.749813-5-gaosong@loongson.cn>

8 days agohw/loongarch: add misc register support dmsi
Song Gao [Tue, 16 Sep 2025 12:21:01 +0000 (20:21 +0800)] 
hw/loongarch: add misc register support dmsi

Add feature register and misc register for dmsi feature checking and
setting

Signed-off-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Message-ID: <20250916122109.749813-4-gaosong@loongson.cn>

8 days agohw/loongarch: add virt feature dmsi support
Song Gao [Tue, 16 Sep 2025 12:21:00 +0000 (20:21 +0800)] 
hw/loongarch: add virt feature dmsi support

dmsi feature is added in LoongArchVirtMachinState, and it is used
to check whether virt machine supports the directy Message-Interrupts.
and by default set dmsi with ON_OFF_AUTO_AUTO.
LoongArchVirtMachineState adds misc_feature and misc_status for misc
features and status. and set the default dintc feature bit.
Msgint feature is added in LoongArchCPU, and it is used to check
whether th cpu supports the Message-Interrupts and by default set
mesgint with ON_OFF_AUTO_AUTO.

Signed-off-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Message-ID: <20250916122109.749813-3-gaosong@loongson.cn>

8 days agotarget/loongarch: move some machine define to virt.h
Song Gao [Tue, 16 Sep 2025 12:20:59 +0000 (20:20 +0800)] 
target/loongarch: move some machine define to virt.h

move some machine define to virt.h

Signed-off-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Bibo Mao <maobibo@loongson.cn>
Message-ID: <20250916122109.749813-2-gaosong@loongson.cn>

8 days agotarget/loongarch: Only flush one TLB entry in helper_invtlb_page_asid()
Bibo Mao [Thu, 4 Sep 2025 11:16:57 +0000 (19:16 +0800)] 
target/loongarch: Only flush one TLB entry in helper_invtlb_page_asid()

With function helper_invtlb_page_asid(), only one TLB entry in
LoongArch emulated TLB is invalidated. so with QEMU TLB, it is not
necessary to flush all QEMU TLB, only flush address range specified
LoongArch emulated TLB is ok. Here invalidate_tlb_entry() is called
so that only QEMU TLB entry with specified address range is flushed.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Only flush one TLB entry in helper_invtlb_page_asid_or_g()
Bibo Mao [Thu, 4 Sep 2025 11:11:25 +0000 (19:11 +0800)] 
target/loongarch: Only flush one TLB entry in helper_invtlb_page_asid_or_g()

With function helper_invtlb_page_asid_or_g(), only one TLB entry in
LoongArch emulated TLB is invalidated. so with QEMU TLB, it is not
necessary to flush all QEMU TLB, only flush address range specified
LoongArch emulated TLB is ok. Here invalidate_tlb_entry() is called
so that only QEMU TLB entry with specified address range is flushed.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Invalid tlb entry in invalidate_tlb()
Bibo Mao [Thu, 4 Sep 2025 10:07:18 +0000 (18:07 +0800)] 
target/loongarch: Invalid tlb entry in invalidate_tlb()

Invalid tlb entry in function invalidate_tlb(), and its usage is
simple and easy to use.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Song Gao <gaosong@loongson.cn>
8 days agotarget/loongarch: Use loongarch_tlb_search_cb in helper_invtlb_page_asid
Bibo Mao [Thu, 4 Sep 2025 09:52:03 +0000 (17:52 +0800)] 
target/loongarch: Use loongarch_tlb_search_cb in helper_invtlb_page_asid

With function helper_invtlb_page_asid(), currently it is to search
TLB entry one by one. Instead STLB can be searched at first with hash
method, and then search MTLB with one by one method

Here common API loongarch_tlb_search_cb() is used in function
helper_invtlb_page_asid()

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Use loongarch_tlb_search_cb in helper_invtlb_page_asid_or_g
Bibo Mao [Thu, 4 Sep 2025 09:46:12 +0000 (17:46 +0800)] 
target/loongarch: Use loongarch_tlb_search_cb in helper_invtlb_page_asid_or_g

With function helper_invtlb_page_asid_or_g(), currently it is to
search TLB entry one by one. Instead STLB can be searched at first
with hash method, and then search MTLB with one by one method.

Here common API loongarch_tlb_search_cb() is used in function
helper_invtlb_page_asid_or_g().

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Change return value type with loongarch_tlb_search_cb()
Bibo Mao [Thu, 4 Sep 2025 09:32:05 +0000 (17:32 +0800)] 
target/loongarch: Change return value type with loongarch_tlb_search_cb()

With function loongarch_tlb_search_cb(), change return value type from
bool type to pointer LoongArchTLB *, the pointer type can be use directly
in future.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Song Gao <gaosong@loongson.cn>
8 days agotarget/loongarch: Add common API loongarch_tlb_search_cb()
Bibo Mao [Thu, 4 Sep 2025 08:06:03 +0000 (16:06 +0800)] 
target/loongarch: Add common API loongarch_tlb_search_cb()

Common API loongarch_tlb_search_cb() is added here to search TLB entry
with specified address.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Add tlb search callback in loongarch_tlb_search()
Bibo Mao [Sat, 2 Aug 2025 02:58:40 +0000 (10:58 +0800)] 
target/loongarch: Add tlb search callback in loongarch_tlb_search()

With function loongarch_tlb_search(), it is to search TLB entry with
speficied virtual address, the difference is selection with asid and
global bit. Here add selection callback with asid and global bit.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Fix page size set issue with CSR_STLBPS
Bibo Mao [Wed, 3 Sep 2025 03:17:56 +0000 (11:17 +0800)] 
target/loongarch: Fix page size set issue with CSR_STLBPS

When modify register CSR_STLBPS, the page size should come from
input parameter rather than old value.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Song Gao <gaosong@loongson.cn>
8 days agotarget/loongarch: Update TLB index selection method
Bibo Mao [Wed, 30 Jul 2025 02:32:54 +0000 (10:32 +0800)] 
target/loongarch: Update TLB index selection method

With function helper_tlbfill(), since there is no suitable TLB entry,
new TLB will be added and flush one old TLB entry. The old TLB entry
index is selected randomly now, instead it can be optimized as
following:
  1. invalid TLB entry can be selected at first.
  2. TLB entry with other ASID can be selected secondly
  3. random method is used by last.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Reduce TLB flush with helper_tlbwr
Bibo Mao [Thu, 24 Jul 2025 12:34:35 +0000 (20:34 +0800)] 
target/loongarch: Reduce TLB flush with helper_tlbwr

With function helper_tlbwr(), specified LoongArch TLB entry will be
updated. There are two PTE pages in one TLB entry called even/odd
pages. Supposing even/odd page is normal/none state, when odd page
is added, TLB entry is changed as normal/normal state and even page
keeps unchanged.

In this situation, it is not necessary to flush QEMU TLB since even
page keep unchanged and odd page is newly changed. Here check whether
PTE page is the same or not, TLB flush can be skipped if both are the
same or newly added.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Add parameter tlb pointer with fill_tlb_entry
Bibo Mao [Thu, 24 Jul 2025 11:57:34 +0000 (19:57 +0800)] 
target/loongarch: Add parameter tlb pointer with fill_tlb_entry

With function fill_tlb_entry(), it will update LoongArch emulated
TLB information. Here parameter tlb pointer is added so that TLB
entry will be updated based on relative TLB CSR registers.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
8 days agotarget/loongarch: Use mmu idx bitmap method when flush TLB
Bibo Mao [Wed, 3 Sep 2025 02:46:01 +0000 (10:46 +0800)] 
target/loongarch: Use mmu idx bitmap method when flush TLB

With API tlb_flush_range_by_mmuidx(), bitmap of mmu idx should be used
rather than itself. Also bitmap of MMU_KERNEL_IDX and MMU_USER_IDX are
used rather than that of current running mmu idx when flush TLB.

Signed-off-by: Bibo Mao <maobibo@loongson.cn>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
9 days agoMerge tag 'pull-target-arm-20250926' of https://gitlab.com/pm215/qemu into staging
Richard Henderson [Fri, 26 Sep 2025 20:27:00 +0000 (13:27 -0700)] 
Merge tag 'pull-target-arm-20250926' of https://gitlab.com/pm215/qemu into staging

target-arm queue:
 * reimplement VHE alias register handling
 * replace magic GIC values by proper definitions
 * convert power control DPRINTF() uses to trace events
 * better reset related tracepoints
 * implement ID_AA64PFR2_EL1
 * hw/usb/hcd-uhci: don't assert for SETUP to non-0 endpoint
 * net/passt: Fix build failure due to missing GIO dependency

# -----BEGIN PGP SIGNATURE-----
#
# iQJNBAABCAA3FiEE4aXFk81BneKOgxXPPCUl7RQ2DN4FAmjWnkUZHHBldGVyLm1h
# eWRlbGxAbGluYXJvLm9yZwAKCRA8JSXtFDYM3upeD/4x0k6ciiJ2wRE1PFUA2KHZ
# bS12+j6Um5BNdcZtRV1aT3x3xOrW3X0JTcmhb9/UdpEPki/krQQgQX50tOiLCeU2
# U4lZke5160Gk3ThdkpELlQDnCVDuNR0wxYgy1GBgAInCa/T/qFnyWwaWBIooCCUh
# +UMJ9tP4XWKvKlkzw9ONFYChxerY2enpOewEbnfSU4NPg9pU8OEZ3yeFWaLZ3Tnl
# 0bei/iFFeuN8RtgJEkuqWI6oENEZZbxGtJ+J/+wvggAfOzfy0I6CmW6y9tQMmKe8
# fTnCQ837uHmlRPWQ615M2wWydbJ1ffdEIYDb5U6UsbfG8sMt5+qg38yo0AyDs6RK
# qJkTceuhqFTDIoi92o2+NFnohCTfASeYaCHjODgcdjGUtbZO7LZ31fOKQrdsHc5e
# chAOnzNxCu9Bt4UqpUmb+ED0fXWDahV1tmgazFS2LORYxnr2q+/WJEdwSgHXNzVy
# 2rdyUx7v7U1finhRE1nAdy8XwJTCQ3gDwDbPGBrH9mhR9DnK6eotFCljI2XnDtAE
# f1i0w/47cnyRW6KsBVK6dJObiOfBRrRYqe3Rt4nA4xjeCNmWcr5IcytpnL/2YT1p
# 1vj+RklbcK7Ns+kWH3H2a9b44zKQrtGGXf8fcNyAqT1YrzrrLUqaiKTfesGfjWit
# ekMWOulOe6UePnoC3SJHFw==
# =+Aj+
# -----END PGP SIGNATURE-----
# gpg: Signature made Fri 26 Sep 2025 07:08:05 AM PDT
# gpg:                using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg:                issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [unknown]
# gpg:                 aka "Peter Maydell <pmaydell@gmail.com>" [unknown]
# gpg:                 aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [unknown]
# gpg:                 aka "Peter Maydell <peter@archaic.org.uk>" [unknown]
# gpg: WARNING: The key's User ID is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* tag 'pull-target-arm-20250926' of https://gitlab.com/pm215/qemu: (44 commits)
  target/arm: Implement ID_AA64PFR2_EL1
  target/arm: Move ID register field defs to cpu-features.h
  target/arm: Trace vCPU reset call
  target/arm: Trace emulated firmware reset call
  target/arm: Convert power control DPRINTF() uses to trace events
  target/arm: Replace magic GIC values by proper definitions
  target/arm: Remove define_arm_vh_e2h_redirects_aliases
  target/arm: Rename some cpreg to their aarch64 names
  target/arm: Redirect VHE FOO_EL12 to FOO_EL1 during translation
  target/arm: Redirect VHE FOO_EL1 -> FOO_EL2 during translation
  target/arm: Split out redirect_cpreg
  target/arm: Rename TBFLAG_A64_NV2_MEM_E20 with *_E2H
  target/arm: Move endianness fixup for 32-bit registers
  target/arm: Move writeback of CP_ANY fields
  target/arm: Move alias setting for wildcards
  target/arm: Remove name argument to alloc_cpreg
  target/arm: Hoist the allocation of ARMCPRegInfo
  target/arm: Split out alloc_cpreg
  target/arm: Add key parameter to add_cpreg_to_hashtable
  target/arm: Move cpreg elimination to define_one_arm_cp_reg
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
9 days agoMerge tag 'pull-10.2-maintainer-260925-1' of https://gitlab.com/stsquad/qemu into...
Richard Henderson [Fri, 26 Sep 2025 20:26:30 +0000 (13:26 -0700)] 
Merge tag 'pull-10.2-maintainer-260925-1' of https://gitlab.com/stsquad/qemu into staging

September maintainer updates (scripts, semihosting, plugins)

 - new gitlab-failure-analysis script
 - tweak checkpath to ignore license in removed lines
 - refactor semihosting to build once
 - add explicit assert to execlog for coverity
 - new uftrace plugin

# -----BEGIN PGP SIGNATURE-----
#
# iQEzBAABCgAdFiEEZoWumedRZ7yvyN81+9DbCVqeKkQFAmjWWJYACgkQ+9DbCVqe
# KkS1sgf+LsP0jsc1wKhzBhO4WarXXacWCDxK22riJ3aolm+gJ+b0WI4ds18A0e3R
# z/J8VJVxBZ+6Hid+tOCQwfZ+Hb1p9IofzBdZryGUvwguviNdlpEChhXXnoZkicym
# aGcC/jYRkhTx42dKRdZrSzPd3ccipqop9RvGx57bjCSBAEHYNz679p4z91kNR5a9
# UfcCzIQHbBUPZo0F9gQkNnBrjsJQhvF+gXPmmsmBI1pby6gNRQvFshrTQ1C32VpL
# VgXNc9cZ6vaREWlgb6izNjsMP7cYTMH2Ppxty/FyEMg7GTfWRjI6Ec8fJKjPFtKr
# ZbCNNAeJ9uLK6pJfTk2YxYabxx3JuQ==
# =cR9e
# -----END PGP SIGNATURE-----
# gpg: Signature made Fri 26 Sep 2025 02:10:46 AM PDT
# gpg:                using RSA key 6685AE99E75167BCAFC8DF35FBD0DB095A9E2A44
# gpg: Good signature from "Alex Bennée (Master Work Key) <alex.bennee@linaro.org>" [unknown]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 6685 AE99 E751 67BC AFC8  DF35 FBD0 DB09 5A9E 2A44

* tag 'pull-10.2-maintainer-260925-1' of https://gitlab.com/stsquad/qemu: (24 commits)
  contrib/plugins/uftrace: add documentation
  contrib/plugins/uftrace_symbols.py
  contrib/plugins/uftrace: implement x64 support
  contrib/plugins/uftrace: generate additional files for uftrace
  contrib/plugins/uftrace: implement privilege level tracing
  contrib/plugins/uftrace: implement tracing
  contrib/plugins/uftrace: track callstack
  contrib/plugins/uftrace: define cpu operations and implement aarch64
  contrib/plugins/uftrace: skeleton file
  contrib/plugins/execlog: Explicitly check for qemu_plugin_read_register() failure
  semihosting/arm-compat-semi: compile once in system and per target for user mode
  semihosting/arm-compat-semi: remove dependency on cpu.h
  semihosting/arm-compat-semi: eradicate target_long
  semihosting/arm-compat-semi: replace target_ulong
  semihosting/arm-compat-semi: eradicate sizeof(target_ulong)
  include/semihosting/common-semi: extract common_semi API
  target/{arm, riscv}/common-semi-target: eradicate target_ulong
  target/riscv/common-semi-target: remove sizeof(target_ulong)
  semihosting/arm-compat-semi: change common_semi_sys_exit_extended
  semihosting/guestfd: compile once for system/user
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
9 days agoMerge tag 'pull-vfio-20250926' of https://github.com/legoater/qemu into staging
Richard Henderson [Fri, 26 Sep 2025 20:26:08 +0000 (13:26 -0700)] 
Merge tag 'pull-vfio-20250926' of https://github.com/legoater/qemu into staging

vfio queue:

* New vfio-user functional test
* Improved naming conventions

# -----BEGIN PGP SIGNATURE-----
#
# iQIzBAABCAAdFiEEoPZlSPBIlev+awtgUaNDx8/77KEFAmjWJaEACgkQUaNDx8/7
# 7KE9Kw/+Ne/AJBxJ0EceK6oCX/1d/56l8dlY7/PHa/LuX8LDzZq42c6+CC2bQ5Ze
# +C5RvM1aavUkvlW73jjmXs1/sdvqq5etGpxbf+hjlLh7uXFaceszWw0+ljH11Pk3
# INZj4iVrPTX7A2d9Gip0ouTFc02DYYllKpUnpbN9MYk0WRgoMBvZISc9T6A28oj1
# Zm24XWOW5ySux9IDIm+zSpvtRp+5r5PgwwwPZvNKBa5jqfJiw6+ZmejLUNEvwfgr
# vkBCGuCywzf+l5wX5z+ofl5GDd6YyRa5W+nDQE7viDwo0XTtZC3WxRKDpqdAruld
# fPjuflqHUpqnyabFuVVc5VV9iyBLGs0cFt2BHPn0vHRZTaCNAGZKZZDzVOP9EYuK
# Jmx5B2vCTnVbg2ZfG+ErEYAVVX97Tgks0xtUxGOpJwbImlRCKdwbrDKSFWT5NYCN
# BlfZX4iFrWLo2fntWP4c3DtM7w6Vn/x5hevyy9pD3i5D8zqDx3FaYJCuM4+5KFAO
# WenxUIkjOcJ829NBfQvF2HtJDEeuLv3eMxszU4pyRnH1vsk38ofOfvZVl2uQd7XT
# NrW207wAAOO+QxrelUGIuuMwhYuh16KnHkjjVe0yEh6htBF5ceIImqZH8SU97qWU
# vQjbcmx/fNQKlKCLc+JOCQz+KJyKahPDGcHi1vxEKXObHFI0rfw=
# =Jl4i
# -----END PGP SIGNATURE-----
# gpg: Signature made Thu 25 Sep 2025 10:33:21 PM PDT
# gpg:                using RSA key A0F66548F04895EBFE6B0B6051A343C7CFFBECA1
# gpg: Good signature from "Cédric Le Goater <clg@redhat.com>" [full]
# gpg:                 aka "Cédric Le Goater <clg@kaod.org>" [full]

* tag 'pull-vfio-20250926' of https://github.com/legoater/qemu: (29 commits)
  include/hw/vfio/vfio-device.h: fix include header guard name
  vfio-user/pci.c: rename vfio_user_pci_dev_info to vfio_user_pci_info
  vfio-user/pci.c: rename vfio_user_instance_finalize() to vfio_user_pci_finalize()
  vfio-user/pci.c: rename vfio_user_instance_init() to vfio_user_pci_init()
  vfio-user/pci.c: rename vfio_user_pci_dev_properties[] to vfio_user_pci_properties[]
  vfio-user/pci.c: rename vfio_user_pci_dev_class_init() to vfio_user_pci_class_init()
  vfio/pci.c: rename vfio_pci_nohotplug_dev_info to vfio_pci_nohotplug_info
  vfio/pci.c: rename vfio_pci_nohotplug_dev_class_init() to vfio_pci_nohotplug_class_init()
  vfio/pci.c: rename vfio_pci_dev_nohotplug_properties[] to vfio_pci_nohotplug_properties[]
  vfio/pci.c: rename vfio_pci_dev_properties[] to vfio_pci_properties[]
  vfio/pci.c: rename vfio_pci_base_dev_info to vfio_pci_device_info
  vfio/pci.c: rename vfio_pci_base_dev_class_init() to vfio_pci_device_class_init()
  hw/vfio/types.h: rename TYPE_VFIO_PCI_BASE to TYPE_VFIO_PCI_DEVICE
  vfio/pci.c: rename vfio_pci_dev_info to vfio_pci_info
  vfio/pci.c: rename vfio_pci_dev_class_init() to vfio_pci_class_init()
  vfio/pci.c: rename vfio_instance_finalize() to vfio_pci_finalize()
  vfio/pci.c: rename vfio_instance_init() to vfio_pci_init()
  vfio/spapr.c: rename VFIOContainer bcontainer field to parent_obj
  vfio/spapr.c: use QOM casts where appropriate
  vfio/vfio-iommufd.h: rename VFIOContainer bcontainer field to parent_obj
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
9 days agotarget/arm: Implement ID_AA64PFR2_EL1
Peter Maydell [Tue, 23 Sep 2025 17:57:51 +0000 (18:57 +0100)] 
target/arm: Implement ID_AA64PFR2_EL1

Currently we define the ID_AA64PFR2_EL1 encoding as reserved (with
the required RAZ behaviour for unassigned system registers in the ID
register encoding space).  Newer architecture versions start to
define fields in this ID register, so define the appropriate
constants and implement it as an ID register backed by a field in
cpu->isar.  Since none of our CPUs set that isar field to non-zero,
there is no behavioural change here (other than the name exposed to
the user via the gdbstub), but this paves the way for implementing
the new features that use fields in this register.

The fields here are the ones documented in rev L.b of the Arm ARM.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
9 days agotarget/arm: Move ID register field defs to cpu-features.h
Peter Maydell [Tue, 23 Sep 2025 17:57:50 +0000 (18:57 +0100)] 
target/arm: Move ID register field defs to cpu-features.h

Currently we define constants for the ID register fields in cpu.h.
This means they're defined for a lot more code in QEMU than actually
needs them.  Move them to cpu-features.h, which is where we define
the feature functions that test fields in these registers.

There's only one place where we need to use some of these macro
definitions that we weren't already including cpu-features.h:
linux-user/arm/target_proc.h.  Otherwise this patch is a pure
movement of code from one file to the other.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
10 days agotarget/arm: Trace vCPU reset call
Philippe Mathieu-Daudé [Wed, 24 Sep 2025 16:32:54 +0000 (18:32 +0200)] 
target/arm: Trace vCPU reset call

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agotarget/arm: Trace emulated firmware reset call
Philippe Mathieu-Daudé [Wed, 24 Sep 2025 16:32:53 +0000 (18:32 +0200)] 
target/arm: Trace emulated firmware reset call

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agotarget/arm: Convert power control DPRINTF() uses to trace events
Philippe Mathieu-Daudé [Wed, 24 Sep 2025 16:32:52 +0000 (18:32 +0200)] 
target/arm: Convert power control DPRINTF() uses to trace events

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agotarget/arm: Replace magic GIC values by proper definitions
Philippe Mathieu-Daudé [Thu, 25 Sep 2025 03:21:51 +0000 (05:21 +0200)] 
target/arm: Replace magic GIC values by proper definitions

Prefer the FIELD_DP64() macro and self-describing GIC
definitions over magic values.

Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agocontrib/plugins/uftrace: add documentation
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:10 +0000 (10:37 +0100)] 
contrib/plugins/uftrace: add documentation

This documentation summarizes how to use the plugin, and present two
examples of the possibilities offered by it, in system and user mode.

As well, it explains how to rebuild and reproduce those examples.

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250902075042.223990-10-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-26-alex.bennee@linaro.org>

10 days agocontrib/plugins/uftrace_symbols.py
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:09 +0000 (10:37 +0100)] 
contrib/plugins/uftrace_symbols.py

usage:  contrib/plugins/uftrace_symbols.py \
        --prefix-symbols \
        arm-trusted-firmware/build/qemu/debug/bl1/bl1.elf \
        arm-trusted-firmware/build/qemu/debug/bl2/bl2.elf \
        arm-trusted-firmware/build/qemu/debug/bl31/bl31.elf \
        u-boot/u-boot:0x60000000 \
        u-boot/u-boot.relocated:0x000000023f6b6000 \
        linux/vmlinux

Will generate symbols and memory mapping files for uftrace, allowing to
have an enhanced trace, instead of raw addresses.

It takes a collection of elf files, and automatically find all their
symbols, and generate an ordered memory map based on that.

This script uses the python (native) pyelftools module.

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Acked-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250902075042.223990-9-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-25-alex.bennee@linaro.org>

10 days agocontrib/plugins/uftrace: implement x64 support
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:08 +0000 (10:37 +0100)] 
contrib/plugins/uftrace: implement x64 support

It's trivial to implement x64 support, as it's the same stack layout
as aarch64.

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250902075042.223990-8-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-24-alex.bennee@linaro.org>

10 days agocontrib/plugins/uftrace: generate additional files for uftrace
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:07 +0000 (10:37 +0100)] 
contrib/plugins/uftrace: generate additional files for uftrace

Beyond traces per cpu, uftrace expect to find some specific files.
- info: contains information about machine/program run
  those values are not impacting uftrace behaviour (only reported by
  uftrace info), and we simply added empty strings.
- memory mapping: how every binary is mapped in memory. For system mode,
  we generate an empty mapping (uftrace_symbols.py, coming in future
  commit, will take care of that). For user mode, we copy current
  /proc/self/maps. We don't need to do any special filtering, as
  reported addresses will necessarily concern guest program, and not
  QEMU and its libraries.
- task: list of tasks. We present every vcpu/privilege level as a
  separate process, as it's the best view we can have when generating a
  (visual) chrome trace. Using threads is less convenient in terms of
  UI.

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250902075042.223990-7-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-23-alex.bennee@linaro.org>

10 days agocontrib/plugins/uftrace: implement privilege level tracing
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:06 +0000 (10:37 +0100)] 
contrib/plugins/uftrace: implement privilege level tracing

We add new option trace-privilege-level=bool, which will create a
separate trace for each privilege level.
This allows to follow changes of privilege during execution.

We implement aarch64 operations to track current privilege level
accordingly.

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250902075042.223990-6-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-22-alex.bennee@linaro.org>

10 days agocontrib/plugins/uftrace: implement tracing
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:05 +0000 (10:37 +0100)] 
contrib/plugins/uftrace: implement tracing

We implement tracing, following uftrace format.
Trace is flushed every 32 MB, so file operations don't impact
performance at runtime.

A different trace is generated per cpu, and we ensure they have a unique
name, based on vcpu_index, while keeping room for privilege level coming
in next commit.

Uftrace format is not officially documented, but it can be found here:
https://github.com/namhyung/uftrace/blob/v0.18/libmcount/record.c#L909

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250902075042.223990-5-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-21-alex.bennee@linaro.org>

10 days agocontrib/plugins/uftrace: track callstack
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:04 +0000 (10:37 +0100)] 
contrib/plugins/uftrace: track callstack

We now track callstack, based on frame pointer analysis. We can detect
function calls, returns, and discontinuities.
We implement a frame pointer based unwinding that is used for
discontinuities.

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250902075042.223990-4-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-20-alex.bennee@linaro.org>

10 days agocontrib/plugins/uftrace: define cpu operations and implement aarch64
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:03 +0000 (10:37 +0100)] 
contrib/plugins/uftrace: define cpu operations and implement aarch64

We define a new CpuOps structure that will be used to implement tracking
independently of guest architecture.

As well, we now instrument only instructions following ones that might
have touched the frame pointer.

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250902075042.223990-3-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-19-alex.bennee@linaro.org>

10 days agocontrib/plugins/uftrace: skeleton file
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:02 +0000 (10:37 +0100)] 
contrib/plugins/uftrace: skeleton file

We define a scoreboard that will hold our data per cpu. As well, we
define a buffer per cpu that will be used to read registers and memories
in a thread-safe way.

For now, we just instrument all instructions with an empty callback.

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250902075042.223990-2-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-18-alex.bennee@linaro.org>

10 days agocontrib/plugins/execlog: Explicitly check for qemu_plugin_read_register() failure
Peter Maydell [Mon, 22 Sep 2025 09:37:01 +0000 (10:37 +0100)] 
contrib/plugins/execlog: Explicitly check for qemu_plugin_read_register() failure

In insn_check_regs() we don't explicitly check whether
qemu_plugin_read_register() failed, which confuses Coverity into
thinking that sz can be -1 in the memcmp().  In fact the assertion
that sz == reg->last->len means this can't happen, but it's clearer
to both humans and Coverity if we explicitly assert that sz > 0, as
we already do in init_vcpu_register().

Coverity: CID 16119011611902
Fixes: af6e4e0a22c1 ("contrib/plugins: extend execlog to track register changes")
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250710144543.1187715-1-peter.maydell@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-17-alex.bennee@linaro.org>

10 days agosemihosting/arm-compat-semi: compile once in system and per target for user mode
Pierrick Bouvier [Mon, 22 Sep 2025 09:37:00 +0000 (10:37 +0100)] 
semihosting/arm-compat-semi: compile once in system and per target for user mode

We don't have any target dependency left in system mode, so we can
compile once.

User mode depends on qemu.h, which is duplicated between linux and bsd,
so we can't easily compile it once.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-13-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-16-alex.bennee@linaro.org>

10 days agosemihosting/arm-compat-semi: remove dependency on cpu.h
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:59 +0000 (10:36 +0100)] 
semihosting/arm-compat-semi: remove dependency on cpu.h

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-12-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-15-alex.bennee@linaro.org>

10 days agosemihosting/arm-compat-semi: eradicate target_long
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:58 +0000 (10:36 +0100)] 
semihosting/arm-compat-semi: eradicate target_long

We use int64_t or int32_t depending on ret size.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-11-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-14-alex.bennee@linaro.org>

10 days agosemihosting/arm-compat-semi: replace target_ulong
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:57 +0000 (10:36 +0100)] 
semihosting/arm-compat-semi: replace target_ulong

Replace with vaddr or uint64_t where appropriate.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-10-pierrick.bouvier@linaro.org>
[AJB: tweak commit message]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-13-alex.bennee@linaro.org>

10 days agosemihosting/arm-compat-semi: eradicate sizeof(target_ulong)
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:56 +0000 (10:36 +0100)] 
semihosting/arm-compat-semi: eradicate sizeof(target_ulong)

No semantic change.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-9-pierrick.bouvier@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-12-alex.bennee@linaro.org>

10 days agoinclude/semihosting/common-semi: extract common_semi API
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:55 +0000 (10:36 +0100)] 
include/semihosting/common-semi: extract common_semi API

We transform target/{arm,riscv}/common-semi-target.h headers to proper
compilation units, and use them in arm-compat-semi.c.

This way, we can include only the declaration header (which is target
agnostic), and selectively link the appropriate implementation based on
current target.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-8-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-11-alex.bennee@linaro.org>

10 days agotarget/{arm, riscv}/common-semi-target: eradicate target_ulong
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:54 +0000 (10:36 +0100)] 
target/{arm, riscv}/common-semi-target: eradicate target_ulong

We replace mechanically with uint64_t.
There is no semantic change, and allows us to extract a proper API from
this set of functions.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-7-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-10-alex.bennee@linaro.org>

10 days agotarget/riscv/common-semi-target: remove sizeof(target_ulong)
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:53 +0000 (10:36 +0100)] 
target/riscv/common-semi-target: remove sizeof(target_ulong)

Only riscv64 extends SYS_EXIT, similar to aarch64.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-6-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-9-alex.bennee@linaro.org>

10 days agosemihosting/arm-compat-semi: change common_semi_sys_exit_extended
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:52 +0000 (10:36 +0100)] 
semihosting/arm-compat-semi: change common_semi_sys_exit_extended

We now check only is sys_exit is extended.
This allows to break dependency to TARGET_SYS_EXIT_EXTENDED which will
not be available anymore from this code.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-5-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-8-alex.bennee@linaro.org>

10 days agosemihosting/guestfd: compile once for system/user
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:51 +0000 (10:36 +0100)] 
semihosting/guestfd: compile once for system/user

We move relevant code to semihosting/arm-compat-semi.c, and add
functions to query CONFIG_ARM_COMPATIBLE_SEMIHOSTING at runtime.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-4-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-7-alex.bennee@linaro.org>

10 days agosemihosting/syscalls: replace uint64_t with vaddr where appropriate
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:50 +0000 (10:36 +0100)] 
semihosting/syscalls: replace uint64_t with vaddr where appropriate

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-3-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-6-alex.bennee@linaro.org>

10 days agosemihosting/syscalls: compile once in system and per target for user mode
Pierrick Bouvier [Mon, 22 Sep 2025 09:36:49 +0000 (10:36 +0100)] 
semihosting/syscalls: compile once in system and per target for user mode

We replace target_ulong mechanically by uint64_t.
We can't compile (easily) this code once for user, as it relies on
various target/function types, so leave it in specific_ss for user mode.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20250822150058.18692-2-pierrick.bouvier@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-5-alex.bennee@linaro.org>

10 days agocheckpatch: Ignore removed lines in license check
Nabih Estefan [Mon, 22 Sep 2025 09:36:48 +0000 (10:36 +0100)] 
checkpatch: Ignore removed lines in license check

When running the license check, if we are updating a license it is
possible for the checkpatch script to test against old license lines
instead of newer ones, since the removal lines appear before the
addition lines in a .patch file.

Fix this by skipping over lines that start with "-" in the checkpatch
script.

Signed-off-by: Nabih Estefan <nabihestefan@google.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250916165928.10048-1-nabihestefan@google.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-4-alex.bennee@linaro.org>

10 days agoscripts/ci: add gitlab-failure-analysis script
Alex Bennée [Mon, 22 Sep 2025 09:36:47 +0000 (10:36 +0100)] 
scripts/ci: add gitlab-failure-analysis script

This is a script designed to collect data from multiple pipelines and
analyse the failure modes they have. By default it will probe the last
3 failed jobs on the staging branch. However this can all be
controlled by the CLI:

  ./scripts/ci/gitlab-failure-analysis --count 2 --branch=testing/next --id 39915562 --status=
  running pipeline 2028486060, total jobs 125, skipped 5, failed 0,  39742 tests, 0 failed tests
  success pipeline 2015018135, total jobs 125, skipped 5, failed 0,  49219 tests, 0 failed tests

You can also skip failing jobs and just dump the tests:

  ./scripts/ci/gitlab-failure-analysis --branch= --id 39915562 --status= --skip-jobs --pipeline 1946202491 1919542960
  failed pipeline 1946202491, total jobs 127, skipped 5, failed 26,  38742 tests, 278 skipped tests, 2 failed tests
    Failed test qemu.qemu:qtest+qtest-s390x / qtest-s390x/boot-serial-test, check-system-opensuse, 1 /s390x/boot-serial/s390-ccw-virtio - FATAL-ERROR: Failed to find expected string. Please check '/tmp/qtest-boot-serial-sW77EA3'
    Failed test qemu.qemu:qtest+qtest-aarch64 / qtest-aarch64/arm-cpu-features, check-system-opensuse, 1 /aarch64/arm/query-cpu-model-expansion - ERROR:../tests/qtest/arm-cpu-features.c:459:test_query_cpu_model_expansion: assertion failed (_error == "The CPU type 'host' requires KVM"): ("The CPU type 'host' requires hardware accelerator" == "The CPU type 'host' requires KVM")
  failed pipeline 1919542960, total jobs 127, skipped 5, failed 2,  48753 tests, 441 skipped tests, 1 failed tests
    Failed test qemu.qemu:unit / test-aio, msys2-64bit, 12 /aio/timer/schedule - ERROR:../tests/unit/test-aio.c:413:test_timer_schedule: assertion failed: (aio_poll(ctx, true))

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20250922093711.2768983-3-alex.bennee@linaro.org>

10 days agoinclude/hw/vfio/vfio-device.h: fix include header guard name
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:36 +0000 (12:31 +0100)] 
include/hw/vfio/vfio-device.h: fix include header guard name

The header guard was incorrectly called HW_VFIO_VFIO_COMMON_H instead of
HW_VFIO_VFIO_DEVICE_H.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-29-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio-user/pci.c: rename vfio_user_pci_dev_info to vfio_user_pci_info
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:35 +0000 (12:31 +0100)] 
vfio-user/pci.c: rename vfio_user_pci_dev_info to vfio_user_pci_info

This changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-28-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio-user/pci.c: rename vfio_user_instance_finalize() to vfio_user_pci_finalize()
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:34 +0000 (12:31 +0100)] 
vfio-user/pci.c: rename vfio_user_instance_finalize() to vfio_user_pci_finalize()

This is the more typical naming convention for QOM finalize() functions, in
particular it changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-27-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio-user/pci.c: rename vfio_user_instance_init() to vfio_user_pci_init()
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:33 +0000 (12:31 +0100)] 
vfio-user/pci.c: rename vfio_user_instance_init() to vfio_user_pci_init()

This is the more typical naming convention for QOM init() functions, in
particular it changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-26-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio-user/pci.c: rename vfio_user_pci_dev_properties[] to vfio_user_pci_properties[]
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:32 +0000 (12:31 +0100)] 
vfio-user/pci.c: rename vfio_user_pci_dev_properties[] to vfio_user_pci_properties[]

This changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-25-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio-user/pci.c: rename vfio_user_pci_dev_class_init() to vfio_user_pci_class_init()
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:31 +0000 (12:31 +0100)] 
vfio-user/pci.c: rename vfio_user_pci_dev_class_init() to vfio_user_pci_class_init()

This changes the function prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-24-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_pci_nohotplug_dev_info to vfio_pci_nohotplug_info
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:30 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_pci_nohotplug_dev_info to vfio_pci_nohotplug_info

This changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-23-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_pci_nohotplug_dev_class_init() to vfio_pci_nohotplug_class_init()
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:29 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_pci_nohotplug_dev_class_init() to vfio_pci_nohotplug_class_init()

This changes the function prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-22-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_pci_dev_nohotplug_properties[] to vfio_pci_nohotplug_properties[]
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:28 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_pci_dev_nohotplug_properties[] to vfio_pci_nohotplug_properties[]

This changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-21-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_pci_dev_properties[] to vfio_pci_properties[]
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:27 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_pci_dev_properties[] to vfio_pci_properties[]

This changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-20-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_pci_base_dev_info to vfio_pci_device_info
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:26 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_pci_base_dev_info to vfio_pci_device_info

This changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-19-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_pci_base_dev_class_init() to vfio_pci_device_class_init()
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:25 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_pci_base_dev_class_init() to vfio_pci_device_class_init()

This changes the function prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-18-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agohw/vfio/types.h: rename TYPE_VFIO_PCI_BASE to TYPE_VFIO_PCI_DEVICE
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:24 +0000 (12:31 +0100)] 
hw/vfio/types.h: rename TYPE_VFIO_PCI_BASE to TYPE_VFIO_PCI_DEVICE

This brings the QOM type name in line with the underlying VFIOPCIDevice structure.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-17-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_pci_dev_info to vfio_pci_info
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:23 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_pci_dev_info to vfio_pci_info

This changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-16-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_pci_dev_class_init() to vfio_pci_class_init()
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:22 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_pci_dev_class_init() to vfio_pci_class_init()

This changes the function prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-15-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_instance_finalize() to vfio_pci_finalize()
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:21 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_instance_finalize() to vfio_pci_finalize()

This is the more typical naming convention for QOM finalize() functions, in
particular it changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-14-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/pci.c: rename vfio_instance_init() to vfio_pci_init()
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:20 +0000 (12:31 +0100)] 
vfio/pci.c: rename vfio_instance_init() to vfio_pci_init()

This is the more typical naming convention for QOM init() functions, in
particular it changes the prefix to match the name of the QOM type.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-13-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/spapr.c: rename VFIOContainer bcontainer field to parent_obj
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:19 +0000 (12:31 +0100)] 
vfio/spapr.c: rename VFIOContainer bcontainer field to parent_obj

Now that nothing accesses the bcontainer field directly, rename bcontainer to
parent_obj as per our current coding guidelines.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-12-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/spapr.c: use QOM casts where appropriate
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:18 +0000 (12:31 +0100)] 
vfio/spapr.c: use QOM casts where appropriate

Use QOM casts to convert between VFIOSpaprContainer and VFIOLegacyContainer
instead of accessing bcontainer directly.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-11-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/vfio-iommufd.h: rename VFIOContainer bcontainer field to parent_obj
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:17 +0000 (12:31 +0100)] 
vfio/vfio-iommufd.h: rename VFIOContainer bcontainer field to parent_obj

Now that nothing accesses the bcontainer field directly, rename bcontainer to
parent_obj as per our current coding guidelines.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-10-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/cpr-iommufd.c: use QOM casts where appropriate
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:16 +0000 (12:31 +0100)] 
vfio/cpr-iommufd.c: use QOM casts where appropriate

Use QOM casts to convert between VFIOIOMMUFDContainer and VFIOContainer instead
of accessing bcontainer directly.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-9-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agovfio/iommufd.c: use QOM casts where appropriate
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:15 +0000 (12:31 +0100)] 
vfio/iommufd.c: use QOM casts where appropriate

Use QOM casts to convert between VFIOIOMMUFDContainer and VFIOContainer instead
of accessing bcontainer directly.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-8-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agohw/vfio/container-base.c: rename file to container.c
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:14 +0000 (12:31 +0100)] 
hw/vfio/container-base.c: rename file to container.c

Rename the file to reflect the previous rename of VFIOContainerBase to
VFIOContainer.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-7-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agohw/vfio/container.c: rename file to container-legacy.c
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:13 +0000 (12:31 +0100)] 
hw/vfio/container.c: rename file to container-legacy.c

This file is mostly concerned with the VFIOLegacyContainer implementation so
rename it to reflect the previous rename of VFIOContainer to
VFIOLegacyContainer.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-6-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agoinclude/hw/vfio/vfio-container-base.h: rename file to vfio-container.h
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:12 +0000 (12:31 +0100)] 
include/hw/vfio/vfio-container-base.h: rename file to vfio-container.h

With the rename of VFIOContainerBase to VFIOContainer, the vfio-container-base.h
header file containing the struct definition is misleading. Rename it from
vfio-container-base.h to vfio-container.h accordingly, fixing up the name
of the include guard at the same time.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-5-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agoinclude/hw/vfio/vfio-container.h: rename file to vfio-container-legacy.h
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:11 +0000 (12:31 +0100)] 
include/hw/vfio/vfio-container.h: rename file to vfio-container-legacy.h

With the rename of VFIOContainer to VFIOLegacyContainer, the vfio-container.h
header file containing the struct definition is misleading. Rename it from
vfio-container.h to vfio-container-legacy.h accordingly, fixing up the name
of the include guard at the same time.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-4-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agoinclude/hw/vfio/vfio-container-base.h: rename VFIOContainerBase to VFIOContainer
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:10 +0000 (12:31 +0100)] 
include/hw/vfio/vfio-container-base.h: rename VFIOContainerBase to VFIOContainer

Now that the VFIOContainer struct name is available, rename VFIOContainerBase
to VFIOContainer to better indicate that it is the superclass of other
VFIOFooContainer structs.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-3-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agoinclude/hw/vfio/vfio-container.h: rename VFIOContainer to VFIOLegacyContainer
Mark Cave-Ayland [Thu, 25 Sep 2025 11:31:09 +0000 (12:31 +0100)] 
include/hw/vfio/vfio-container.h: rename VFIOContainer to VFIOLegacyContainer

The VFIOContainer struct represents the legacy VFIO container even though the
name suggests it may be the common superclass of all VFIO containers. Rename it
to VFIOLegacyContainer to make this clearer, which is also a better match for
its VFIO_IOMMU_LEGACY QOM type name.

Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250925113159.1760317-2-mark.caveayland@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agotests/functional: add a vfio-user smoke test
Mark Cave-Ayland [Thu, 11 Sep 2025 21:09:05 +0000 (23:09 +0200)] 
tests/functional: add a vfio-user smoke test

Add a basic test of the vfio-user PCI client implementation.

Co-authored-by: John Levon <john.levon@nutanix.com>
Signed-off-by: Mark Cave-Ayland <mark.caveayland@nutanix.com>
Signed-off-by: John Levon <john.levon@nutanix.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250911210905.2070474-1-john.levon@nutanix.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
10 days agotarget/arm: Remove define_arm_vh_e2h_redirects_aliases
Richard Henderson [Tue, 16 Sep 2025 14:22:37 +0000 (07:22 -0700)] 
target/arm: Remove define_arm_vh_e2h_redirects_aliases

Populate vhe_redir_to_{el2,el01} on each ARMCPRegInfo.
Clear the fields within add_cpreg_to_hashtable_aa32.
Create the FOO_EL12 cpreg within add_cpreg_to_hashtable_aa64;
add ARM_CP_NO_RAW.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agotarget/arm: Rename some cpreg to their aarch64 names
Richard Henderson [Tue, 16 Sep 2025 14:22:36 +0000 (07:22 -0700)] 
target/arm: Rename some cpreg to their aarch64 names

Rename those registers which will have FOO_EL12 aliases.

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agotarget/arm: Redirect VHE FOO_EL12 to FOO_EL1 during translation
Richard Henderson [Tue, 16 Sep 2025 14:22:35 +0000 (07:22 -0700)] 
target/arm: Redirect VHE FOO_EL12 to FOO_EL1 during translation

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
[PMM: expanded a comment slightly]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agotarget/arm: Redirect VHE FOO_EL1 -> FOO_EL2 during translation
Richard Henderson [Tue, 16 Sep 2025 14:22:34 +0000 (07:22 -0700)] 
target/arm: Redirect VHE FOO_EL1 -> FOO_EL2 during translation

Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agotarget/arm: Split out redirect_cpreg
Richard Henderson [Tue, 16 Sep 2025 14:22:33 +0000 (07:22 -0700)] 
target/arm: Split out redirect_cpreg

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
10 days agotarget/arm: Rename TBFLAG_A64_NV2_MEM_E20 with *_E2H
Richard Henderson [Tue, 16 Sep 2025 14:22:32 +0000 (07:22 -0700)] 
target/arm: Rename TBFLAG_A64_NV2_MEM_E20 with *_E2H

Install e2h in tbflags and compute nv2_mem_e20 from
that in aarch64_tr_init_disas_context.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>