]> git.proxmox.com Git - mirror_qemu.git/log
mirror_qemu.git
2 years agohw/intc: sifive_plic: Move the properties
Alistair Francis [Mon, 18 Oct 2021 02:39:11 +0000 (12:39 +1000)]
hw/intc: sifive_plic: Move the properties

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Message-id: 3c125e27c49a4969df82bf8b197535ccd1996939.1634524691.git.alistair.francis@wdc.com

2 years agohw/intc: Remove the Ibex PLIC
Alistair Francis [Mon, 18 Oct 2021 02:38:55 +0000 (12:38 +1000)]
hw/intc: Remove the Ibex PLIC

The Ibex PLIC is now spec compliant. Let's remove the Ibex PLIC and
instead use the SiFive PLIC.

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Message-id: 5557935c2660c5e6281b6d21e6514e019593662e.1634524691.git.alistair.francis@wdc.com

2 years agohw/riscv: opentitan: Update to the latest build
Alistair Francis [Mon, 18 Oct 2021 02:38:39 +0000 (12:38 +1000)]
hw/riscv: opentitan: Update to the latest build

Update the OpenTitan machine model to match the latest OpenTitan FPGA
design.

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Message-id: 18b1b681b0f8dd2461e819d1217bf0b530812680.1634524691.git.alistair.francis@wdc.com

2 years agotarget/riscv: Compute mstatus.sd on demand
Richard Henderson [Wed, 20 Oct 2021 03:17:09 +0000 (20:17 -0700)]
target/riscv: Compute mstatus.sd on demand

The position of this read-only field is dependent on the current xlen.
Rather than having to compute that difference in many places, compute
it only on read.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-16-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Use riscv_csrrw_debug for cpu_dump
Richard Henderson [Wed, 20 Oct 2021 03:17:08 +0000 (20:17 -0700)]
target/riscv: Use riscv_csrrw_debug for cpu_dump

Use the official debug read interface to the csrs,
rather than referencing the env slots directly.
Put the list of csrs to dump into a table.

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-15-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Use gen_shift*_per_ol for RVB, RVI
Richard Henderson [Wed, 20 Oct 2021 03:17:07 +0000 (20:17 -0700)]
target/riscv: Use gen_shift*_per_ol for RVB, RVI

Most shift instructions require a separate implementation
for RV32 when TARGET_LONG_BITS == 64.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-14-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Use gen_unary_per_ol for RVB
Richard Henderson [Wed, 20 Oct 2021 03:17:06 +0000 (20:17 -0700)]
target/riscv: Use gen_unary_per_ol for RVB

The count zeros instructions require a separate implementation
for RV32 when TARGET_LONG_BITS == 64.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-13-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Adjust trans_rev8_32 for riscv64
Richard Henderson [Wed, 20 Oct 2021 03:17:05 +0000 (20:17 -0700)]
target/riscv: Adjust trans_rev8_32 for riscv64

When target_long is 64-bit, we still want a 32-bit bswap for rev8.
Since this opcode is specific to RV32, we need not conditionalize.

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-12-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Use gen_arith_per_ol for RVM
Richard Henderson [Wed, 20 Oct 2021 03:17:04 +0000 (20:17 -0700)]
target/riscv: Use gen_arith_per_ol for RVM

The multiply high-part instructions require a separate
implementation for RV32 when TARGET_LONG_BITS == 64.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-11-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Replace DisasContext.w with DisasContext.ol
Richard Henderson [Wed, 20 Oct 2021 03:17:03 +0000 (20:17 -0700)]
target/riscv: Replace DisasContext.w with DisasContext.ol

In preparation for RV128, consider more than just "w" for
operand size modification.  This will be used for the "d"
insns from RV128 as well.

Rename oper_len to get_olen to better match get_xlen.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-10-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Replace is_32bit with get_xl/get_xlen
Richard Henderson [Wed, 20 Oct 2021 03:17:02 +0000 (20:17 -0700)]
target/riscv: Replace is_32bit with get_xl/get_xlen

In preparation for RV128, replace a simple predicate
with a more versatile test.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-9-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Properly check SEW in amo_op
Richard Henderson [Wed, 20 Oct 2021 03:17:01 +0000 (20:17 -0700)]
target/riscv: Properly check SEW in amo_op

We're currently assuming SEW <= 3, and the "else" from
the SEW == 3 must be less.  Use a switch and explicitly
bound both SEW and SEQ for all cases.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-8-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Use REQUIRE_64BIT in amo_check64
Richard Henderson [Wed, 20 Oct 2021 03:17:00 +0000 (20:17 -0700)]
target/riscv: Use REQUIRE_64BIT in amo_check64

Use the same REQUIRE_64BIT check that we use elsewhere,
rather than open-coding the use of is_32bit.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-7-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Add MXL/SXL/UXL to TB_FLAGS
Richard Henderson [Wed, 20 Oct 2021 03:16:59 +0000 (20:16 -0700)]
target/riscv: Add MXL/SXL/UXL to TB_FLAGS

Begin adding support for switching XLEN at runtime.  Extract the
effective XLEN from MISA and MSTATUS and store for use during translation.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-6-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Replace riscv_cpu_is_32bit with riscv_cpu_mxl
Richard Henderson [Wed, 20 Oct 2021 03:16:58 +0000 (20:16 -0700)]
target/riscv: Replace riscv_cpu_is_32bit with riscv_cpu_mxl

Shortly, the set of supported XL will not be just 32 and 64,
and representing that properly using the enumeration will be
imperative.

Two places, booting and gdb, intentionally use misa_mxl_max
to emphasize the use of the reset value of misa.mxl, and not
the current cpu state.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-5-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Split misa.mxl and misa.ext
Richard Henderson [Wed, 20 Oct 2021 03:16:57 +0000 (20:16 -0700)]
target/riscv: Split misa.mxl and misa.ext

The hw representation of misa.mxl is at the high bits of the
misa csr.  Representing this in the same way inside QEMU
results in overly complex code trying to check that field.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-4-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Create RISCVMXL enumeration
Richard Henderson [Wed, 20 Oct 2021 03:16:56 +0000 (20:16 -0700)]
target/riscv: Create RISCVMXL enumeration

Move the MXL_RV* defines to enumerators.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-3-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Move cpu_get_tb_cpu_state out of line
Richard Henderson [Wed, 20 Oct 2021 03:16:55 +0000 (20:16 -0700)]
target/riscv: Move cpu_get_tb_cpu_state out of line

Move the function to cpu_helper.c, as it is large and growing.

Reviewed-by: LIU Zhiwei <zhiwei_liu@c-sky.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211020031709.359469-2-richard.henderson@linaro.org
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Organise the CPU properties
Alistair Francis [Mon, 18 Oct 2021 04:32:15 +0000 (14:32 +1000)]
target/riscv: Organise the CPU properties

Organise the CPU properties so that standard extensions come first
then followed by experimental extensions.

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Message-id: b6598570f60c5ee7f402be56d837bb44b289cc4d.1634531504.git.alistair.francis@wdc.com

2 years agotarget/riscv: Remove some unused macros
Alistair Francis [Mon, 18 Oct 2021 04:32:00 +0000 (14:32 +1000)]
target/riscv: Remove some unused macros

Since commit 1a9540d1f1a
("target/riscv: Drop support for ISA spec version 1.09.1")
these definitions are unused, remove them.

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Message-id: f4d8a7a035f39c0a35d44c1e371c5c99cc2fa15a.1634531504.git.alistair.francis@wdc.com

2 years agotarget/riscv: fix TB_FLAGS bits overlapping bug for rvv/rvh
Frank Chang [Fri, 15 Oct 2021 07:45:02 +0000 (15:45 +0800)]
target/riscv: fix TB_FLAGS bits overlapping bug for rvv/rvh

TB_FLAGS mem_idx bits was extended from 2 bits to 3 bits in
commit: c445593, but other TB_FLAGS bits for rvv and rvh were
not shift as well so these bits may overlap with each other when
rvv is enabled.

Signed-off-by: Frank Chang <frank.chang@sifive.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Message-Id: <20211015074627.3957162-2-frank.chang@sifive.com>
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agohw/riscv: virt: Use machine->ram as the system memory
Mingwang Li [Sat, 16 Oct 2021 03:09:08 +0000 (11:09 +0800)]
hw/riscv: virt: Use machine->ram as the system memory

If default main_mem is used to be registered as the system memory,
other memory cannot be initialized. Therefore, the system memory
should be initialized to the machine->ram, which consists of the
default main_mem and other possible memory required by applications,
such as shared hugepage memory in DPDK.

Also, the mc->defaul_ram_id should be set to the default main_mem,
such as "riscv_virt_board.ram" for the virt machine.

Signed-off-by: Mingwang Li <limingwang@huawei.com>
Signed-off-by: Yifei Jiang <jiangyifei@huawei.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Message-id: 20211016030908.40480-1-limingwang@huawei.com
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Fix orc.b implementation
Philipp Tomsich [Wed, 13 Oct 2021 18:41:25 +0000 (20:41 +0200)]
target/riscv: Fix orc.b implementation

The earlier implementation fell into a corner case for bytes that were
0x01, giving a wrong result (but not affecting our application test
cases for strings, as an ASCII value 0x01 is rare in those...).

This changes the algorithm to:
 1. Mask out the high-bit of each bytes (so that each byte is <= 127).
 2. Add 127 to each byte (i.e. if the low 7 bits are not 0, this will overflow
    into the highest bit of each byte).
 3. Bitwise-or the original value back in (to cover those cases where the
    source byte was exactly 128) to saturate the high-bit.
 4. Shift-and-mask (implemented as a mask-and-shift) to extract the MSB of
    each byte into its LSB.
 5. Multiply with 0xff to fan out the LSB to all bits of each byte.

Fixes: d7a4fcb034 ("target/riscv: Add orc.b instruction for Zbb, removing gorc/gorci")
Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu>
Reported-by: Vincent Palatin <vpalatin@rivosinc.com>
Tested-by: Vincent Palatin <vpalatin@rivosinc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211013184125.2010897-1-philipp.tomsich@vrull.eu
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: line up all of the registers in the info register dump
Travis Geiselbrecht [Sat, 9 Oct 2021 05:50:19 +0000 (22:50 -0700)]
target/riscv: line up all of the registers in the info register dump

Ensure the columns for all of the register names and values line up.
No functional change, just a minor tweak to the output.

Signed-off-by: Travis Geiselbrecht <travisg@gmail.com>
Reviewed-by: Alistair Francis <alistair.francis@wdc.com>
Message-id: 20211009055019.545153-1-travisg@gmail.com
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agotarget/riscv: Pass the same value to oprsz and maxsz for vmv.v.v
Frank Chang [Thu, 7 Oct 2021 08:17:41 +0000 (16:17 +0800)]
target/riscv: Pass the same value to oprsz and maxsz for vmv.v.v

oprsz and maxsz are passed with the same value in commit: eee2d61e202.
However, vmv.v.v was missed in that commit and should pass the same
value as well in its tcg_gen_gvec_2_ptr() call.

Signed-off-by: Frank Chang <frank.chang@sifive.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20211007081803.1705656-1-frank.chang@sifive.com
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
2 years agoMerge remote-tracking branch 'remotes/rth/tags/pull-arm-20211021' into staging
Richard Henderson [Thu, 21 Oct 2021 16:53:27 +0000 (09:53 -0700)]
Merge remote-tracking branch 'remotes/rth/tags/pull-arm-20211021' into staging

Introduce cpu topology support
Generate DBG2 table
Switch to ssize_t for elf loader return type
Fixed sbsa cpu type error message typo
Only initialize required submodules for edk2
Dont create device-tree node for empty NUMA node

# gpg: Signature made Thu 21 Oct 2021 08:22:32 AM PDT
# gpg:                using RSA key 7A481E78868B4DB6A85A05C064DF38E8AF7E215F
# gpg:                issuer "richard.henderson@linaro.org"
# gpg: Good signature from "Richard Henderson <richard.henderson@linaro.org>" [ultimate]

* remotes/rth/tags/pull-arm-20211021:
  tests/data/acpi/virt: Update the empty expected file for PPTT
  hw/arm/virt-acpi-build: Generate PPTT table
  tests/data/acpi/virt: Add an empty expected file for PPTT
  hw/acpi/aml-build: Add PPTT table
  hw/acpi/aml-build: Add Processor hierarchy node structure
  hw/arm/virt: Add cpu-map to device tree
  device_tree: Add qemu_fdt_add_path
  hw/arm/virt: Only describe cpu topology since virt-6.2
  bios-tables-test: Generate reference table for virt/DBG2
  hw/arm/virt_acpi_build: Generate DBG2 table
  tests/acpi: Add void table for virt/DBG2 bios-tables-test
  hw/elf_ops.h: switch to ssize_t for elf loader return type
  hw/arm/sbsa-ref: Fixed cpu type error message typo.
  roms/edk2: Only initialize required submodules
  roms/edk2: Only init brotli submodule to build BaseTools
  hw/arm/virt: Don't create device-tree node for empty NUMA node
  tests/acpi: Generate reference blob for IORT rev E.b
  hw/arm/virt-acpi-build: IORT upgrade up to revision E.b
  tests/acpi: Get prepared for IORT E.b revision upgrade

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agoMerge remote-tracking branch 'remotes/dgibson/tags/ppc-for-6.2-20211021' into staging
Richard Henderson [Thu, 21 Oct 2021 15:27:46 +0000 (08:27 -0700)]
Merge remote-tracking branch 'remotes/dgibson/tags/ppc-for-6.2-20211021' into staging

ppc patch queue 2021-10-21

Here's the next batch of ppc target related patches for qemu-6.2.
Highlights are:
 * Some fixes and minimal tests for old embedded ppc platforms
 * The beginnings of PMU emulation in TCG from Daniel Barboza
 * Some improvements to the pegasos2 platform
 * A number of TCG bugfixes from the folks at the El Dorado Institute
 * A few other assorted bugfixes and cleanups

# gpg: Signature made Wed 20 Oct 2021 09:19:04 PM PDT
# gpg:                using RSA key 75F46586AE61A66CC44E87DC6C38CACA20D9B392
# gpg: Good signature from "David Gibson <david@gibson.dropbear.id.au>" [full]
# gpg:                 aka "David Gibson (kernel.org) <dwg@kernel.org>" [unknown]
# gpg:                 aka "David Gibson (Red Hat) <dgibson@redhat.com>" [full]
# gpg:                 aka "David Gibson (ozlabs.org) <dgibson@ozlabs.org>" [full]

* remotes/dgibson/tags/ppc-for-6.2-20211021: (25 commits)
  hw/ppc/ppc4xx_pci: Fix ppc4xx_pci_map_irq() for recent Linux kernels
  target/ppc: adding user read/write functions for PMCs
  target/ppc: add user read/write functions for MMCR2
  target/ppc: add user read/write functions for MMCR0
  target/ppc: add MMCR0 PMCC bits to hflags
  target/ppc: Filter mtmsr[d] input before setting MSR
  tests/acceptance: Add a test for the bamboo ppc board
  ppc/pegasos2: Implement power-off RTAS function with VOF
  ppc/pegasos2: Add constants for PCI config addresses
  ppc/pegasos2: Access MV64361 registers via their memory region
  ppc/pegasos2: Implement get-time-of-day RTAS function with VOF
  ppc/pegasos2: Warn when using VOF but no kernel is specified
  ppc/pegasos2: Restrict memory to 2 gigabytes
  target/ppc: Fix XER access in monitor
  linux-user: Fix XER access in ppc version of elf_core_copy_regs
  target/ppc: Fix XER access in gdbstub
  linux-user/ppc: Fix XER access in save/restore_user_regs
  tests/acceptance: Add tests for the ppc405 boards
  hw/ppc: Fix iothread locking in the 405 code
  spapr/xive: Use xive_esb_rw() to trigger interrupts
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agotests/data/acpi/virt: Update the empty expected file for PPTT
Yanan Wang [Wed, 20 Oct 2021 14:21:25 +0000 (22:21 +0800)]
tests/data/acpi/virt: Update the empty expected file for PPTT

Run ./tests/data/acpi/rebuild-expected-aml.sh from build directory
to update PPTT binary. Also empty bios-tables-test-allowed-diff.h.

Disassembled output of the updated new file:

/*
 * Intel ACPI Component Architecture
 * AML/ASL+ Disassembler version 20180810 (64-bit version)
 * Copyright (c) 2000 - 2018 Intel Corporation
 *
 * Disassembly of tests/data/acpi/virt/PPTT, Fri Oct  8 10:12:32 2021
 *
 * ACPI Data Table [PPTT]
 *
 * Format: [HexOffset DecimalOffset ByteLength]  FieldName : FieldValue
 */

[000h 0000   4]                    Signature : "PPTT"    [Processor Properties Topology Table]
[004h 0004   4]                 Table Length : 0000004C
[008h 0008   1]                     Revision : 02
[009h 0009   1]                     Checksum : A8
[00Ah 0010   6]                       Oem ID : "BOCHS "
[010h 0016   8]                 Oem Table ID : "BXPC    "
[018h 0024   4]                 Oem Revision : 00000001
[01Ch 0028   4]              Asl Compiler ID : "BXPC"
[020h 0032   4]        Asl Compiler Revision : 00000001

[024h 0036   1]                Subtable Type : 00 [Processor Hierarchy Node]
[025h 0037   1]                       Length : 14
[026h 0038   2]                     Reserved : 0000
[028h 0040   4]        Flags (decoded below) : 00000001
                            Physical package : 1
                     ACPI Processor ID valid : 0
[02Ch 0044   4]                       Parent : 00000000
[030h 0048   4]            ACPI Processor ID : 00000000
[034h 0052   4]      Private Resource Number : 00000000

[038h 0056   1]                Subtable Type : 00 [Processor Hierarchy Node]
[039h 0057   1]                       Length : 14
[03Ah 0058   2]                     Reserved : 0000
[03Ch 0060   4]        Flags (decoded below) : 0000000A
                            Physical package : 0
                     ACPI Processor ID valid : 1
[040h 0064   4]                       Parent : 00000024
[044h 0068   4]            ACPI Processor ID : 00000000
[048h 0072   4]      Private Resource Number : 00000000

Raw Table Data: Length 76 (0x4C)

    0000: 50 50 54 54 4C 00 00 00 02 A8 42 4F 43 48 53 20  // PPTTL.....BOCHS
    0010: 42 58 50 43 20 20 20 20 01 00 00 00 42 58 50 43  // BXPC    ....BXPC
    0020: 01 00 00 00 00 14 00 00 01 00 00 00 00 00 00 00  // ................
    0030: 00 00 00 00 00 00 00 00 00 14 00 00 0A 00 00 00  // ................
    0040: 24 00 00 00 00 00 00 00 00 00 00 00              // $...........

Reviewed-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Message-Id: <20211020142125.7516-9-wangyanan55@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/arm/virt-acpi-build: Generate PPTT table
Yanan Wang [Wed, 20 Oct 2021 14:21:24 +0000 (22:21 +0800)]
hw/arm/virt-acpi-build: Generate PPTT table

Generate the Processor Properties Topology Table (PPTT) for ARM
virt machines supporting it (>= 6.2).

Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Message-Id: <20211020142125.7516-8-wangyanan55@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agotests/data/acpi/virt: Add an empty expected file for PPTT
Yanan Wang [Wed, 20 Oct 2021 14:21:23 +0000 (22:21 +0800)]
tests/data/acpi/virt: Add an empty expected file for PPTT

Add a generic empty binary file for the new introduced PPTT table
under tests/data/acpi/virt, and list it as files to be changed in
tests/qtest/bios-tables-test-allowed-diff.h

Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Message-Id: <20211020142125.7516-7-wangyanan55@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/acpi/aml-build: Add PPTT table
Andrew Jones [Wed, 20 Oct 2021 14:21:22 +0000 (22:21 +0800)]
hw/acpi/aml-build: Add PPTT table

Add the Processor Properties Topology Table (PPTT) used to
describe CPU topology information to ACPI guests.

Note, a DT-boot Linux guest with a non-flat CPU topology will
see socket and core IDs being sequential integers starting
from zero, which is different from ACPI-boot Linux guest,
e.g. with -smp 4,sockets=2,cores=2,threads=1

a DT boot produces:

 cpu:  0 package_id:  0 core_id:  0
 cpu:  1 package_id:  0 core_id:  1
 cpu:  2 package_id:  1 core_id:  0
 cpu:  3 package_id:  1 core_id:  1

an ACPI boot produces:

 cpu:  0 package_id: 36 core_id:  0
 cpu:  1 package_id: 36 core_id:  1
 cpu:  2 package_id: 96 core_id:  2
 cpu:  3 package_id: 96 core_id:  3

This is due to several reasons:

 1) DT cpu nodes do not have an equivalent field to what the PPTT
    ACPI Processor ID must be, i.e. something equal to the MADT CPU
    UID or equal to the UID of an ACPI processor container. In both
    ACPI cases those are platform dependant IDs assigned by the
    vendor.

 2) While QEMU is the vendor for a guest, if the topology specifies
    SMT (> 1 thread), then, with ACPI, it is impossible to assign a
    core-id the same value as a package-id, thus it is not possible
    to have package-id=0 and core-id=0. This is because package and
    core containers must be in the same ACPI namespace and therefore
    must have unique UIDs.

 3) ACPI processor containers are not mandatorily required for PPTT
    tables to be used and, due to the limitations of which IDs are
    selected described above in (2), they are not helpful for QEMU,
    so we don't build them with this patch. In the absence of them,
    Linux assigns its own unique IDs. The maintainers have chosen not
    to use counters from zero, but rather ACPI table offsets, which
    explains why the numbers are so much larger than with DT.

 4) When there is no SMT (threads=1) the core IDs for ACPI boot guests
    match the logical CPU IDs, because these IDs must be equal to the
    MADT CPU UID (as no processor containers are present), and QEMU
    uses the logical CPU ID for these MADT IDs.

So in summary, with QEMU as the vendor for the guests, we simply
use sequential integers starting from zero for the non-leaf nodes
but with ID-valid flag unset, so that guest will ignore them and
use table offsets as unique container IDs. And we use logical CPU
IDs for the leaf nodes with the ID-valid flag set, which will be
consistent with MADT.

Currently the implementation of PPTT generation complies with ACPI
specification 5.2.29 (Revision 6.3). The 6.3 spec can be found at:
https://uefi.org/sites/default/files/resources/ACPI_6_3_May16.pdf

Reviewed-by: Eric Auger <eric.auger@redhat.com>
Co-developed-by: Yanan Wang <wangyanan55@huawei.com>
Signed-off-by: Andrew Jones <drjones@redhat.com>
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20211020142125.7516-6-wangyanan55@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/acpi/aml-build: Add Processor hierarchy node structure
Yanan Wang [Wed, 20 Oct 2021 14:21:21 +0000 (22:21 +0800)]
hw/acpi/aml-build: Add Processor hierarchy node structure

Add a generic API to build Processor hierarchy node structure (Type 0),
which is strictly consistent with descriptions in ACPI 6.3: 5.2.29.1.

This function will be used to build ACPI PPTT table for cpu topology.

Co-developed-by: Ying Fang <fangying1@huawei.com>
Co-developed-by: Henglong Fan <fanhenglong@huawei.com>
Co-developed-by: Yanan Wang <wangyanan55@huawei.com>
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Message-Id: <20211020142125.7516-5-wangyanan55@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/arm/virt: Add cpu-map to device tree
Andrew Jones [Wed, 20 Oct 2021 14:21:20 +0000 (22:21 +0800)]
hw/arm/virt: Add cpu-map to device tree

Support device tree CPU topology descriptions.

In accordance with the Devicetree Specification, the Linux Doc
"arm/cpus.yaml" requires that cpus and cpu nodes in the DT are
present. And we have already met the requirement by generating
/cpus/cpu@* nodes for members within ms->smp.cpus. Accordingly,
we should also create subnodes in cpu-map for the present cpus,
each of which relates to an unique cpu node.

The Linux Doc "cpu/cpu-topology.txt" states that the hierarchy
of CPUs in a SMP system is defined through four entities and
they are socket/cluster/core/thread. It is also required that
a socket node's child nodes must be one or more cluster nodes.
Given that currently we are only provided with information of
socket/core/thread, we assume there is one cluster child node
in each socket node when creating cpu-map.

Co-developed-by: Yanan Wang <wangyanan55@huawei.com>
Signed-off-by: Andrew Jones <drjones@redhat.com>
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Message-Id: <20211020142125.7516-4-wangyanan55@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agodevice_tree: Add qemu_fdt_add_path
Yanan Wang [Wed, 20 Oct 2021 14:21:19 +0000 (22:21 +0800)]
device_tree: Add qemu_fdt_add_path

qemu_fdt_add_path() works like qemu_fdt_add_subnode(), except it
also adds all missing subnodes from the given path. We'll use it
in a coming patch where we will add cpu-map to the device tree.

And we also tweak an error message of qemu_fdt_add_subnode().

Co-developed-by: Andrew Jones <drjones@redhat.com>
Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: Alistair Francis <alistair.francis@wdc.com>
Message-Id: <20211020142125.7516-3-wangyanan55@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/arm/virt: Only describe cpu topology since virt-6.2
Yanan Wang [Wed, 20 Oct 2021 14:21:18 +0000 (22:21 +0800)]
hw/arm/virt: Only describe cpu topology since virt-6.2

On existing older machine types, without cpu topology described
in ACPI or DT, the guest will populate one by default. With the
topology described, it will read the information and set up its
topology as instructed, but that may not be the same as what was
getting used by default. It's possible that an user application
has a dependency on the default topology and if the default one
gets changed it will probably behave differently.

Based on above consideration we'd better only describe topology
information to the guest on 6.2 and later machine types.

Signed-off-by: Yanan Wang <wangyanan55@huawei.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <20211020142125.7516-2-wangyanan55@huawei.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agobios-tables-test: Generate reference table for virt/DBG2
Eric Auger [Tue, 19 Oct 2021 08:00:37 +0000 (10:00 +0200)]
bios-tables-test: Generate reference table for virt/DBG2

Add the DBG2 table generated with
tests/data/acpi/rebuild-expected-aml.sh

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20211019080037.930641-4-eric.auger@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/arm/virt_acpi_build: Generate DBG2 table
Eric Auger [Tue, 19 Oct 2021 08:00:36 +0000 (10:00 +0200)]
hw/arm/virt_acpi_build: Generate DBG2 table

ARM SBBR specification mandates DBG2 table (Debug Port Table 2)
since v1.0 (ARM DEN0044F 8.3.1.7 DBG2).

The DBG2 table allows to describe one or more debug ports.

Generate an DBG2 table featuring a single debug port, the PL011.

The DBG2 specification can be found at
"Microsoft Debug Port Table 2 (DBG2)"
https://docs.microsoft.com/en-us/windows-hardware/drivers/bringup/acpi-debug-port-table?redirectedfrom=MSDN

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20211019080037.930641-3-eric.auger@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agotests/acpi: Add void table for virt/DBG2 bios-tables-test
Eric Auger [Tue, 19 Oct 2021 08:00:35 +0000 (10:00 +0200)]
tests/acpi: Add void table for virt/DBG2 bios-tables-test

Add placeholders for DBG2 reference table for
virt tests and ignore till reference blob is added.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Acked-by: Igor Mammedov <imammedo@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20211019080037.930641-2-eric.auger@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/ppc/ppc4xx_pci: Fix ppc4xx_pci_map_irq() for recent Linux kernels
Thomas Huth [Tue, 19 Oct 2021 09:18:17 +0000 (11:18 +0200)]
hw/ppc/ppc4xx_pci: Fix ppc4xx_pci_map_irq() for recent Linux kernels

Recent Linux kernels are accessing the PCI device in slot 0 that
represents the PCI host bridge. This causes ppc4xx_pci_map_irq()
to return -1 which causes an assert() later:

 hw/pci/pci.c:262: pci_bus_change_irq_level: Assertion `irq_num >= 0' failed.

Thus we should allocate an IRQ line for the device in slot 0, too.
To avoid changes to the outside of ppc4xx_pci.c, we map it to
the internal IRQ number 4 which will then happily be ignored since
ppc440_bamboo.c does not wire it up.

With these changes it is now possible again to use recent Linux
kernels for the bamboo board.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20211019091817.469003-1-thuth@redhat.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Tested-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: adding user read/write functions for PMCs
Daniel Henrique Barboza [Mon, 18 Oct 2021 01:01:22 +0000 (22:01 -0300)]
target/ppc: adding user read/write functions for PMCs

Problem state needs to be able to read and write the PMU counters,
otherwise it won't be aware of any sampling result that the PMU produces
after a Perf run.

This patch does that in a similar fashion as already done in the
previous patches. PMCs 5 and 6 have a special condition, aside from the
constraints that are common with PMCs 1-4, where they are not part of the
PMU if MMCR0_PMCC is 0b11.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20211018010133.315842-5-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: add user read/write functions for MMCR2
Daniel Henrique Barboza [Mon, 18 Oct 2021 01:01:21 +0000 (22:01 -0300)]
target/ppc: add user read/write functions for MMCR2

Similar to the previous patch, let's add problem state read/write access to
the MMCR2 SPR, which is also a group A PMU SPR that needs to be filtered
to be read/written by userspace.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20211018010133.315842-4-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: add user read/write functions for MMCR0
Gustavo Romero [Mon, 18 Oct 2021 01:01:20 +0000 (22:01 -0300)]
target/ppc: add user read/write functions for MMCR0

Userspace need access to PMU SPRs to be able to operate the PMU. One of
such SPRs is MMCR0.

MMCR0, as defined by PowerISA v3.1, is classified as a 'group A' PMU
register. This class of registers has common read/write rules that are
governed by MMCR0 PMCC bits. MMCR0 is also not fully exposed to problem
state: only MMCR0_FC, MMCR0_PMAO and MMCR0_PMAE bits are
readable/writable in this case.

This patch exposes MMCR0 to userspace by doing the following:

- two new callbacks, spr_read_MMCR0_ureg() and spr_write_MMCR0_ureg(),
are added to be used as problem state read/write callbacks of UMMCR0.
Both callbacks filters the amount of bits userspace is able to
read/write by using a MMCR0_UREG_MASK;

- problem state access control is done by the spr_groupA_read_allowed()
and spr_groupA_write_allowed() helpers. These helpers will read the
current PMCC bits from DisasContext and check whether the read/write
MMCR0 operation is valid or noti;

- to avoid putting exclusive PMU logic into the already loaded
translate.c file, let's create a new 'power8-pmu-regs.c.inc' file that
will hold all the spr_read/spr_write functions of PMU registers.

The 'power8' name of this new file intends to hint about the proven
support of the PMU logic to be added. The code has been tested with the
IBM POWER chip family, POWER8 being the oldest version tested. This
doesn't mean that the PMU logic will break with any other PPC64 chip
that implements Book3s, but rather that we can't assert that it works
properly with any Book3s compliant chip.

CC: Gustavo Romero <gustavo.romero@linaro.org>
Signed-off-by: Gustavo Romero <gromero@linux.ibm.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20211018010133.315842-3-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: add MMCR0 PMCC bits to hflags
Daniel Henrique Barboza [Mon, 18 Oct 2021 01:01:19 +0000 (22:01 -0300)]
target/ppc: add MMCR0 PMCC bits to hflags

We're going to add PMU support for TCG PPC64 chips, based on IBM POWER8+
emulation and following PowerISA v3.1. This requires several PMU related
registers to be exposed to userspace (problem state). PowerISA v3.1
dictates that the PMCC bits of the MMCR0 register controls the level of
access of the PMU registers to problem state.

This patch start things off by exposing both PMCC bits to hflags,
allowing us to access them via DisasContext in the read/write callbacks
that we're going to add next.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20211018010133.315842-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: Filter mtmsr[d] input before setting MSR
Matheus Ferst [Fri, 15 Oct 2021 18:19:40 +0000 (15:19 -0300)]
target/ppc: Filter mtmsr[d] input before setting MSR

PowerISA says that mtmsr[d] "does not alter MSR[HV], MSR[S], MSR[ME], or
MSR[LE]", but the current code only filters the GPR-provided value if
L=1. This behavior caused some problems in FreeBSD, and a build option
was added to work around the issue [1], but it seems that the bug was
not reported in launchpad/gitlab. This patch address the issue in qemu,
so the option on FreeBSD should no longer be required.

[1] https://cgit.freebsd.org/src/commit/?id=4efb1ca7d2a44cfb33d7f9e18bd92f8d68dcfee0

Signed-off-by: Matheus Ferst <matheus.ferst@eldorado.org.br>
Message-Id: <20211015181940.197982-1-matheus.ferst@eldorado.org.br>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotests/acceptance: Add a test for the bamboo ppc board
Thomas Huth [Fri, 15 Oct 2021 09:00:08 +0000 (11:00 +0200)]
tests/acceptance: Add a test for the bamboo ppc board

The kernel and initrd from the "Aboriginal Linux" project can be
used to run some tests on the bamboo ppc machine.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20211015090008.1299609-1-thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agoppc/pegasos2: Implement power-off RTAS function with VOF
BALATON Zoltan [Thu, 14 Oct 2021 19:50:19 +0000 (21:50 +0200)]
ppc/pegasos2: Implement power-off RTAS function with VOF

This only helps Linux guests as only that seems to use it.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Message-Id: <1c1e030f2bbc86e950b3310fb5922facdc21ef86.1634241019.git.balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agoppc/pegasos2: Add constants for PCI config addresses
BALATON Zoltan [Thu, 14 Oct 2021 19:50:19 +0000 (21:50 +0200)]
ppc/pegasos2: Add constants for PCI config addresses

Define a constant for PCI config addresses to make it clearer what
these numbers are.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Message-Id: <9bd8e84d02d91693b71082a1fadeb86e6bce3025.1634241019.git.balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agoppc/pegasos2: Access MV64361 registers via their memory region
BALATON Zoltan [Thu, 14 Oct 2021 19:50:19 +0000 (21:50 +0200)]
ppc/pegasos2: Access MV64361 registers via their memory region

Instead of relying on the mapped address of the MV64361 registers
access them via their memory region. This is not a problem at reset
time when these registers are mapped at the default address but the
guest could change this later and then the RTAS calls accessing PCI
config registers could fail. None of the guests actually do this so
this only avoids a theoretical problem not seen in practice.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Message-Id: <b6f768023603dc2c4d130720bcecdbea459b7668.1634241019.git.balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agoppc/pegasos2: Implement get-time-of-day RTAS function with VOF
BALATON Zoltan [Thu, 14 Oct 2021 19:50:19 +0000 (21:50 +0200)]
ppc/pegasos2: Implement get-time-of-day RTAS function with VOF

This is needed for Linux to access RTC time.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Message-Id: <6233eb07c680d6c74427e11b9641958f98d53378.1634241019.git.balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agoppc/pegasos2: Warn when using VOF but no kernel is specified
BALATON Zoltan [Thu, 14 Oct 2021 19:50:19 +0000 (21:50 +0200)]
ppc/pegasos2: Warn when using VOF but no kernel is specified

Issue a warning when using VOF (which is the default) but no -kernel
option given to let users know that it will likely fail as the guest
has nothing to run. It is not a hard error because it may still be
useful to start the machine without further options for testing or
inspecting it from monitor without actually booting it.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Message-Id: <a4ec9a900df772b91e9f69ca7a0799d8ae293e5a.1634241019.git.balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agoppc/pegasos2: Restrict memory to 2 gigabytes
BALATON Zoltan [Thu, 14 Oct 2021 19:50:19 +0000 (21:50 +0200)]
ppc/pegasos2: Restrict memory to 2 gigabytes

The CHRP spec this board confirms to only allows 2 GiB of system
memory below 4 GiB as the high 2 GiB is allocated to IO and system
resources. To avoid problems with memory overlapping these areas
restrict RAM to 2 GiB similar to mac_newworld.

Signed-off-by: BALATON Zoltan <balaton@eik.bme.hu>
Message-Id: <54f58229a69c9c1cca21bcecad700b3d7052edd5.1634241019.git.balaton@eik.bme.hu>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: Fix XER access in monitor
Matheus Ferst [Thu, 14 Oct 2021 22:32:34 +0000 (19:32 -0300)]
target/ppc: Fix XER access in monitor

We can't read env->xer directly, as it does not contain some bits of
XER. Instead, we should have a callback that uses cpu_read_xer to read
the complete register.

Fixes: da91a00f191f ("target-ppc: Split out SO, OV, CA fields from XER")
Signed-off-by: Matheus Ferst <matheus.ferst@eldorado.org.br>
Message-Id: <20211014223234.127012-5-matheus.ferst@eldorado.org.br>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agolinux-user: Fix XER access in ppc version of elf_core_copy_regs
Matheus Ferst [Thu, 14 Oct 2021 22:32:33 +0000 (19:32 -0300)]
linux-user: Fix XER access in ppc version of elf_core_copy_regs

env->xer doesn't hold some bits of XER, like OV and CA. To write the
complete register in the core dump we should read XER value with
cpu_read_xer.

Reported-by: Lucas Mateus Castro (alqotel) <lucas.araujo@eldorado.org.br>
Fixes: da91a00f191f ("target-ppc: Split out SO, OV, CA fields from XER")
Signed-off-by: Matheus Ferst <matheus.ferst@eldorado.org.br>
Message-Id: <20211014223234.127012-4-matheus.ferst@eldorado.org.br>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: Fix XER access in gdbstub
Matheus Ferst [Thu, 14 Oct 2021 22:32:32 +0000 (19:32 -0300)]
target/ppc: Fix XER access in gdbstub

The value of XER is split in multiple fields of CPUPPCState, like
env->xer and env->so. To get/set the whole register from gdb, we should
use cpu_read_xer/cpu_write_xer.

Fixes: da91a00f191f ("target-ppc: Split out SO, OV, CA fields from XER")
Signed-off-by: Matheus Ferst <matheus.ferst@eldorado.org.br>
Message-Id: <20211014223234.127012-3-matheus.ferst@eldorado.org.br>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agolinux-user/ppc: Fix XER access in save/restore_user_regs
Matheus Ferst [Thu, 14 Oct 2021 22:32:31 +0000 (19:32 -0300)]
linux-user/ppc: Fix XER access in save/restore_user_regs

We should use cpu_read_xer/cpu_write_xer to save/restore the complete
register since some of its bits are in other fields of CPUPPCState. A
test is added to prevent future regressions.

Fixes: da91a00f191f ("target-ppc: Split out SO, OV, CA fields from XER")
Signed-off-by: Matheus Ferst <matheus.ferst@eldorado.org.br>
Message-Id: <20211014223234.127012-2-matheus.ferst@eldorado.org.br>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotests/acceptance: Add tests for the ppc405 boards
Thomas Huth [Mon, 11 Oct 2021 12:59:30 +0000 (14:59 +0200)]
tests/acceptance: Add tests for the ppc405 boards

Using the U-Boot firmware, we can check that at least the serial console
of the ppc405 boards is still usable.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20211011125930.750217-1-thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
[dwg: Added an extra tag at Philippe's suggestion]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agohw/ppc: Fix iothread locking in the 405 code
Thomas Huth [Wed, 6 Oct 2021 07:11:40 +0000 (09:11 +0200)]
hw/ppc: Fix iothread locking in the 405 code

When using u-boot as firmware with the taihu board, QEMU aborts with
this assertion:

 ERROR:../accel/tcg/tcg-accel-ops.c:79:tcg_handle_interrupt: assertion failed:
  (qemu_mutex_iothread_locked())

Running QEMU with "-d in_asm" shows that the crash happens when writing
to SPR 0x3f2, so we are missing to lock the iothread in the code path
here.

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20211006071140.565952-1-thuth@redhat.com>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Tested-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agospapr/xive: Use xive_esb_rw() to trigger interrupts
Cédric Le Goater [Wed, 6 Oct 2021 21:05:46 +0000 (23:05 +0200)]
spapr/xive: Use xive_esb_rw() to trigger interrupts

xive_esb_rw() is the common routine used for memory accesses on ESB
page. Use it for triggers also.

Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20211006210546.641102-1-clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agohw/ppc/spapr_softmmu: Reduce include list
Philippe Mathieu-Daudé [Wed, 6 Oct 2021 17:08:01 +0000 (19:08 +0200)]
hw/ppc/spapr_softmmu: Reduce include list

Commit 962104f0448 ("hw/ppc: moved hcalls that depend on softmmu")
introduced a lot of unnecessary #include directives. Remove them.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Message-Id: <20211006170801.178023-1-philmd@redhat.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: Fix the test raising the decrementer exception
Cédric Le Goater [Tue, 5 Oct 2021 05:33:24 +0000 (07:33 +0200)]
target/ppc: Fix the test raising the decrementer exception

Commit 4d9b8ef9b5ab ("target/ppc: Fix 64-bit decrementer") introduced
new int64t variables and broke the test triggering the decrementer
exception. Revert partially the change to evaluate both clause of the
if statement.

Reported-by: Coverity CID 1464061
Fixes: 4d9b8ef9b5ab ("target/ppc: Fix 64-bit decrementer")
Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20211005053324.441132-1-clg@kaod.org>
Reviewed-by: Greg Kurz <groug@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: Use tcg_constant_i64() in gen_brh()
Philippe Mathieu-Daudé [Sun, 3 Oct 2021 14:17:11 +0000 (16:17 +0200)]
target/ppc: Use tcg_constant_i64() in gen_brh()

The mask of the Byte-Reverse Halfword opcode is a read-only
constant. We can avoid using a TCG temporary by moving the
mask to the constant pool.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20211003141711.3673181-3-f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agotarget/ppc: Use tcg_constant_i32() in gen_setb()
Philippe Mathieu-Daudé [Sun, 3 Oct 2021 14:17:10 +0000 (16:17 +0200)]
target/ppc: Use tcg_constant_i32() in gen_setb()

Avoid using TCG temporaries for the -1 and 8 constant values.

Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20211003141711.3673181-2-f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agospapr/xive: Add source status helpers
Cédric Le Goater [Mon, 4 Oct 2021 21:21:41 +0000 (23:21 +0200)]
spapr/xive: Add source status helpers

and use them to set and test the ASSERTED bit of LSI sources.

Signed-off-by: Cédric Le Goater <clg@kaod.org>
Message-Id: <20211004212141.432954-1-clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2 years agohw/elf_ops.h: switch to ssize_t for elf loader return type
Luc Michel [Thu, 14 Oct 2021 19:43:25 +0000 (21:43 +0200)]
hw/elf_ops.h: switch to ssize_t for elf loader return type

Until now, int was used as the return type for all the ELF
loader related functions. The returned value is the sum of all loaded
program headers "MemSize" fields.

Because of the overflow check in elf_ops.h, trying to load an ELF bigger
than INT_MAX will fail. Switch to ssize_t to remove this limitation.

Signed-off-by: Luc Michel <lmichel@kalray.eu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-Id: <20211014194325.19917-1-lmichel@kalray.eu>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/arm/sbsa-ref: Fixed cpu type error message typo.
Shuuichirou Ishii [Fri, 8 Oct 2021 06:36:04 +0000 (15:36 +0900)]
hw/arm/sbsa-ref: Fixed cpu type error message typo.

Signed-off-by: Shuuichirou Ishii <ishii.shuuichir@fujitsu.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Yanan Wang <wangyanan55@huawei.com>
Message-Id: <20211008063604.670699-1-ishii.shuuichir@fujitsu.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agoroms/edk2: Only initialize required submodules
Philippe Mathieu-Daudé [Mon, 18 Oct 2021 10:58:16 +0000 (12:58 +0200)]
roms/edk2: Only initialize required submodules

The EDK2 firmware images built to test QEMU do not require
the following submodules:

 - MdeModulePkg/Universal/RegularExpressionDxe/oniguruma
 - UnitTestFrameworkPkg/Library/CmockaLib/cmocka

The only submodules required are:

 - ArmPkg/Library/ArmSoftFloatLib/berkeley-softfloat-3
 - BaseTools/Source/C/BrotliCompress/brotli
 - CryptoPkg/Library/OpensslLib/openssl
 - MdeModulePkg/Library/BrotliCustomDecompressLib/brotli

Adapt the buildsys machinery to only initialize the required
submodules.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>
Message-Id: <20211018105816.2663195-3-philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agoroms/edk2: Only init brotli submodule to build BaseTools
Philippe Mathieu-Daudé [Mon, 18 Oct 2021 10:58:15 +0000 (12:58 +0200)]
roms/edk2: Only init brotli submodule to build BaseTools

Since EDK2 BaseTools only require the brotli submodule,
we don't need to initialize other submodules to build it.

Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Gerd Hoffmann <kraxel@redhat.com>
Message-Id: <20211018105816.2663195-2-philmd@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/arm/virt: Don't create device-tree node for empty NUMA node
Gavin Shan [Fri, 15 Oct 2021 12:42:46 +0000 (20:42 +0800)]
hw/arm/virt: Don't create device-tree node for empty NUMA node

The empty NUMA node, where no memory resides, are allowed. For
example, the following command line specifies two empty NUMA nodes.
With this, QEMU fails to boot because of the conflicting device-tree
node names, as the following error message indicates.

  /home/gavin/sandbox/qemu.main/build/qemu-system-aarch64 \
  -accel kvm -machine virt,gic-version=host               \
  -cpu host -smp 4,sockets=2,cores=2,threads=1            \
  -m 1024M,slots=16,maxmem=64G                            \
  -object memory-backend-ram,id=mem0,size=512M            \
  -object memory-backend-ram,id=mem1,size=512M            \
  -numa node,nodeid=0,cpus=0-1,memdev=mem0                \
  -numa node,nodeid=1,cpus=2-3,memdev=mem1                \
  -numa node,nodeid=2                                     \
  -numa node,nodeid=3
    :
  qemu-system-aarch64: FDT: Failed to create subnode /memory@80000000: FDT_ERR_EXISTS

As specified by linux device-tree binding document, the device-tree
nodes for these empty NUMA nodes shouldn't be generated. However,
the corresponding NUMA node IDs should be included in the distance
map. The memory hotplug through device-tree on ARM64 isn't existing
so far and it's not necessary to require the user to provide a distance
map. Furthermore, the default distance map Linux generates may even be
sufficient. So this simply skips populating the device-tree nodes for
these empty NUMA nodes to avoid the error, so that QEMU can be started
successfully.

Signed-off-by: Gavin Shan <gshan@redhat.com>
Reviewed-by: Andrew Jones <drjones@redhat.com>
Message-Id: <20211015124246.23073-1-gshan@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agotests/acpi: Generate reference blob for IORT rev E.b
Eric Auger [Thu, 14 Oct 2021 11:56:43 +0000 (13:56 +0200)]
tests/acpi: Generate reference blob for IORT rev E.b

Re-generate reference blobs with rebuild-expected-aml.sh.

Differences reported by "make check V=1" are listed below
(IORT.numamem). Differences for other variants are similar.

 /*
  * Intel ACPI Component Architecture
  * AML/ASL+ Disassembler version 20180629 (64-bit version)
  * Copyright (c) 2000 - 2018 Intel Corporation
  *
- * Disassembly of tests/data/acpi/virt/IORT.numamem, Thu Oct 14 06:13:19 2021
+ * Disassembly of /tmp/aml-K8L9A1, Thu Oct 14 06:13:19 2021
  *
  * ACPI Data Table [IORT]
  *
  * Format: [HexOffset DecimalOffset ByteLength]  FieldName : FieldValue
  */

 [000h 0000   4]                    Signature : "IORT"    [IO Remapping Table]
-[004h 0004   4]                 Table Length : 0000007C
-[008h 0008   1]                     Revision : 00
-[009h 0009   1]                     Checksum : 07
+[004h 0004   4]                 Table Length : 00000080
+[008h 0008   1]                     Revision : 03
+[009h 0009   1]                     Checksum : B3
 [00Ah 0010   6]                       Oem ID : "BOCHS "
 [010h 0016   8]                 Oem Table ID : "BXPC    "
 [018h 0024   4]                 Oem Revision : 00000001
 [01Ch 0028   4]              Asl Compiler ID : "BXPC"
 [020h 0032   4]        Asl Compiler Revision : 00000001

 [024h 0036   4]                   Node Count : 00000002
 [028h 0040   4]                  Node Offset : 00000030
 [02Ch 0044   4]                     Reserved : 00000000

 [030h 0048   1]                         Type : 00
 [031h 0049   2]                       Length : 0018
-[033h 0051   1]                     Revision : 00
+[033h 0051   1]                     Revision : 01
 [034h 0052   4]                     Reserved : 00000000
 [038h 0056   4]                Mapping Count : 00000000
 [03Ch 0060   4]               Mapping Offset : 00000000

 [040h 0064   4]                     ItsCount : 00000001
 [044h 0068   4]                  Identifiers : 00000000

 [048h 0072   1]                         Type : 02
-[049h 0073   2]                       Length : 0034
-[04Bh 0075   1]                     Revision : 00
-[04Ch 0076   4]                     Reserved : 00000000
+[049h 0073   2]                       Length : 0038
+[04Bh 0075   1]                     Revision : 03
+[04Ch 0076   4]                     Reserved : 00000001
 [050h 0080   4]                Mapping Count : 00000001
-[054h 0084   4]               Mapping Offset : 00000020
+[054h 0084   4]               Mapping Offset : 00000024

 [058h 0088   8]            Memory Properties : [IORT Memory Access Properties]
 [058h 0088   4]              Cache Coherency : 00000001
 [05Ch 0092   1]        Hints (decoded below) : 00
                                    Transient : 0
                               Write Allocate : 0
                                Read Allocate : 0
                                     Override : 0
 [05Dh 0093   2]                     Reserved : 0000
 [05Fh 0095   1] Memory Flags (decoded below) : 03
                                    Coherency : 1
                             Device Attribute : 1
 [060h 0096   4]                ATS Attribute : 00000000
 [064h 0100   4]           PCI Segment Number : 00000000
-[068h 0104   1]            Memory Size Limit : 00
+[068h 0104   1]            Memory Size Limit : 40
 [069h 0105   3]                     Reserved : 000000

-[068h 0104   4]                   Input base : 00000000
-[06Ch 0108   4]                     ID Count : 0000FFFF
-[070h 0112   4]                  Output Base : 00000000
-[074h 0116   4]             Output Reference : 00000030
-[078h 0120   4]        Flags (decoded below) : 00000000
+[06Ch 0108   4]                   Input base : 00000000
+[070h 0112   4]                     ID Count : 0000FFFF
+[074h 0116   4]                  Output Base : 00000000
+[078h 0120   4]             Output Reference : 00000030
+[07Ch 0124   4]        Flags (decoded below) : 00000000
                               Single Mapping : 0

-Raw Table Data: Length 124 (0x7C)
+Raw Table Data: Length 128 (0x80)

-    0000: 49 4F 52 54 7C 00 00 00 00 07 42 4F 43 48 53 20  // IORT|.....BOCHS
+    0000: 49 4F 52 54 80 00 00 00 03 B3 42 4F 43 48 53 20  // IORT......BOCHS
     0010: 42 58 50 43 20 20 20 20 01 00 00 00 42 58 50 43  // BXPC    ....BXPC
     0020: 01 00 00 00 02 00 00 00 30 00 00 00 00 00 00 00  // ........0.......
-    0030: 00 18 00 00 00 00 00 00 00 00 00 00 00 00 00 00  // ................
-    0040: 01 00 00 00 00 00 00 00 02 34 00 00 00 00 00 00  // .........4......
-    0050: 01 00 00 00 20 00 00 00 01 00 00 00 00 00 00 03  // .... ...........
-    0060: 00 00 00 00 00 00 00 00 00 00 00 00 FF FF 00 00  // ................
-    0070: 00 00 00 00 30 00 00 00 00 00 00 00              // ....0.......
+    0030: 00 18 00 01 00 00 00 00 00 00 00 00 00 00 00 00  // ................
+    0040: 01 00 00 00 00 00 00 00 02 38 00 03 01 00 00 00  // .........8......
+    0050: 01 00 00 00 24 00 00 00 01 00 00 00 00 00 00 03  // ....$...........
+    0060: 00 00 00 00 00 00 00 00 40 00 00 00 00 00 00 00  // ........@.......
+    0070: FF FF 00 00 00 00 00 00 30 00 00 00 00 00 00 00  // ........0.......
**

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20211014115643.756977-4-eric.auger@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agohw/arm/virt-acpi-build: IORT upgrade up to revision E.b
Eric Auger [Thu, 14 Oct 2021 11:56:42 +0000 (13:56 +0200)]
hw/arm/virt-acpi-build: IORT upgrade up to revision E.b

Upgrade the IORT table from B to E.b specification
revision (ARM DEN 0049E.b).

The SMMUv3 and root complex node have additional
fields. Also unique IORT node identifiers are
introduced: they are generated in sequential order.
They are not cross-referenced though.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20211014115643.756977-3-eric.auger@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agotests/acpi: Get prepared for IORT E.b revision upgrade
Eric Auger [Thu, 14 Oct 2021 11:56:41 +0000 (13:56 +0200)]
tests/acpi: Get prepared for IORT E.b revision upgrade

Ignore IORT till reference blob for E.b spec revision gets
added.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-Id: <20211014115643.756977-2-eric.auger@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agoMerge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging
Richard Henderson [Wed, 20 Oct 2021 13:10:51 +0000 (06:10 -0700)]
Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging

pc,pci,virtio: features, fixes, tests

vhost user rng
vdpa multiqueue
Fixes, cleanups, new tests all over the place.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Wed 20 Oct 2021 03:18:24 AM PDT
# gpg:                using RSA key 5D09FD0871C8F85B94CA8A0D281F0DB8D28D5469
# gpg:                issuer "mst@redhat.com"
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>" [full]
# gpg:                 aka "Michael S. Tsirkin <mst@redhat.com>" [full]

* remotes/mst/tags/for_upstream: (44 commits)
  tests/acpi/bios-tables-test: update DSDT blob for multifunction bridge test
  tests/acpi/pcihp: add unit tests for hotplug on multifunction bridges for q35
  tests/acpi/bios-tables-test: add and allow changes to a new q35 DSDT table blob
  pci: fix PCI resource reserve capability on BE
  vhost-vdpa: multiqueue support
  virtio-net: vhost control virtqueue support
  vhost: record the last virtqueue index for the virtio device
  virtio-net: use "queue_pairs" instead of "queues" when possible
  vhost-net: control virtqueue support
  net: introduce control client
  vhost-vdpa: let net_vhost_vdpa_init() returns NetClientState *
  vhost-vdpa: prepare for the multiqueue support
  vhost-vdpa: classify one time request
  vhost-vdpa: open device fd in net_init_vhost_vdpa()
  bios-tables-test: don't disassemble empty files
  rebuild-expected-aml.sh: allow partial target list
  qdev/qbus: remove failover specific code
  vhost-user-blk-test: pass vhost-user socket fds to QSD
  failover: fix a regression introduced by JSON'ification of -device
  vhost-user: fix duplicated notifier MR init
  ...

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2 years agotests/acpi/bios-tables-test: update DSDT blob for multifunction bridge test
Ani Sinha [Thu, 7 Oct 2021 13:57:50 +0000 (19:27 +0530)]
tests/acpi/bios-tables-test: update DSDT blob for multifunction bridge test

We added a new unit test for testing acpi hotplug on multifunction bridges in
q35 machines. Here, we update the DSDT table gloden master blob for this unit
test.

The test adds the following devices to qemu and then checks the changes
introduced in the DSDT table due to the addition of the following devices:

(a) a multifunction bridge device
(b) a bridge device with function 1
(c) a non-bridge device with function 2

In the DSDT table, we should see AML hotplug descriptions for (a) and (b).
For (a) we should find a hotplug AML description for function 0.

Following is the ASL diff between the original DSDT table and the modified DSDT
table due to the unit test. We see that multifunction bridge on bus 2 and single
function bridge on bus 3 function 1 are described, not the non-bridge balloon
device on bus 4, function 2.

@@ -1,30 +1,30 @@
 /*
  * Intel ACPI Component Architecture
  * AML/ASL+ Disassembler version 20190509 (64-bit version)
  * Copyright (c) 2000 - 2019 Intel Corporation
  *
  * Disassembling to symbolic ASL+ operators
  *
- * Disassembly of tests/data/acpi/q35/DSDT, Thu Oct  7 18:29:19 2021
+ * Disassembly of /tmp/aml-C7JCA1, Thu Oct  7 18:29:19 2021
  *
  * Original Table Header:
  *     Signature        "DSDT"
- *     Length           0x00002061 (8289)
+ *     Length           0x00002187 (8583)
  *     Revision         0x01 **** 32-bit table (V1), no 64-bit math support
- *     Checksum         0xF9
+ *     Checksum         0x8D
  *     OEM ID           "BOCHS "
  *     OEM Table ID     "BXPC    "
  *     OEM Revision     0x00000001 (1)
  *     Compiler ID      "BXPC"
  *     Compiler Version 0x00000001 (1)
  */
 DefinitionBlock ("", "DSDT", 1, "BOCHS ", "BXPC    ", 0x00000001)
 {
     Scope (\)
     {
         OperationRegion (DBG, SystemIO, 0x0402, One)
         Field (DBG, ByteAcc, NoLock, Preserve)
         {
             DBGB,   8
         }

@@ -3265,23 +3265,95 @@
                 Method (_S1D, 0, NotSerialized)  // _S1D: S1 Device State
                 {
                     Return (Zero)
                 }

                 Method (_S2D, 0, NotSerialized)  // _S2D: S2 Device State
                 {
                     Return (Zero)
                 }

                 Method (_S3D, 0, NotSerialized)  // _S3D: S3 Device State
                 {
                     Return (Zero)
                 }
             }

+            Device (S10)
+            {
+                Name (_ADR, 0x00020000)  // _ADR: Address
+                Name (BSEL, One)
+                Device (S00)
+                {
+                    Name (_SUN, Zero)  // _SUN: Slot User Number
+                    Name (_ADR, Zero)  // _ADR: Address
+                    Method (_EJ0, 1, NotSerialized)  // _EJx: Eject Device, x=0-9
+                    {
+                        PCEJ (BSEL, _SUN)
+                    }
+
+                    Method (_DSM, 4, Serialized)  // _DSM: Device-Specific Method
+                    {
+                        Return (PDSM (Arg0, Arg1, Arg2, Arg3, BSEL, _SUN))
+                    }
+                }
+
+                Method (DVNT, 2, NotSerialized)
+                {
+                    If ((Arg0 & One))
+                    {
+                        Notify (S00, Arg1)
+                    }
+                }
+
+                Method (PCNT, 0, NotSerialized)
+                {
+                    BNUM = One
+                    DVNT (PCIU, One)
+                    DVNT (PCID, 0x03)
+                }
+            }
+
+            Device (S19)
+            {
+                Name (_ADR, 0x00030001)  // _ADR: Address
+                Name (BSEL, Zero)
+                Device (S00)
+                {
+                    Name (_SUN, Zero)  // _SUN: Slot User Number
+                    Name (_ADR, Zero)  // _ADR: Address
+                    Method (_EJ0, 1, NotSerialized)  // _EJx: Eject Device, x=0-9
+                    {
+                        PCEJ (BSEL, _SUN)
+                    }
+
+                    Method (_DSM, 4, Serialized)  // _DSM: Device-Specific Method
+                    {
+                        Return (PDSM (Arg0, Arg1, Arg2, Arg3, BSEL, _SUN))
+                    }
+                }
+
+                Method (DVNT, 2, NotSerialized)
+                {
+                    If ((Arg0 & One))
+                    {
+                        Notify (S00, Arg1)
+                    }
+                }
+
+                Method (PCNT, 0, NotSerialized)
+                {
+                    BNUM = Zero
+                    DVNT (PCIU, One)
+                    DVNT (PCID, 0x03)
+                }
+            }
+
             Method (PCNT, 0, NotSerialized)
             {
+                ^S19.PCNT ()
+                ^S10.PCNT ()
             }
         }
     }
 }

Signed-off-by: Ani Sinha <ani@anisinha.ca>
Message-Id: <20211007135750.1277213-4-ani@anisinha.ca>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Igor Mammedov <imammedo@redhat.com>
2 years agotests/acpi/pcihp: add unit tests for hotplug on multifunction bridges for q35
Ani Sinha [Thu, 7 Oct 2021 13:57:49 +0000 (19:27 +0530)]
tests/acpi/pcihp: add unit tests for hotplug on multifunction bridges for q35

commit d7346e614f4ec ("acpi: x86: pcihp: add support hotplug on multifunction bridges")
added ACPI hotplug descriptions for cold plugged bridges for functions other
than 0. For all other devices, the ACPI hotplug descriptions are limited to
function 0 only. This change adds unit tests for this feature.

This test adds the following devices to qemu and then checks the changes
introduced in the DSDT table due to the addition of the following devices:

(a) a multifunction bridge device
(b) a bridge device with function 1
(c) a non-bridge device with function 2

In the DSDT table, we should see AML hotplug descriptions for (a) and (b).
For (a) we should find a hotplug AML description for function 0.

The following diff compares the DSDT table AML with the new unit test before
and after the change d7346e614f4ec is introduced. In other words,
this diff reflects the changes that occurs in the DSDT table due to the change
d7346e614f4ec .

@@ -1,60 +1,38 @@
 /*
  * Intel ACPI Component Architecture
  * AML/ASL+ Disassembler version 20190509 (64-bit version)
  * Copyright (c) 2000 - 2019 Intel Corporation
  *
  * Disassembling to symbolic ASL+ operators
  *
- * Disassembly of tests/data/acpi/q35/DSDT.multi-bridge, Thu Oct  7 18:56:05 2021
+ * Disassembly of /tmp/aml-AN0DA1, Thu Oct  7 18:56:05 2021
  *
  * Original Table Header:
  *     Signature        "DSDT"
- *     Length           0x000020FE (8446)
+ *     Length           0x00002187 (8583)
  *     Revision         0x01 **** 32-bit table (V1), no 64-bit math support
- *     Checksum         0xDE
+ *     Checksum         0x8D
  *     OEM ID           "BOCHS "
  *     OEM Table ID     "BXPC    "
  *     OEM Revision     0x00000001 (1)
  *     Compiler ID      "BXPC"
  *     Compiler Version 0x00000001 (1)
  */
 DefinitionBlock ("", "DSDT", 1, "BOCHS ", "BXPC    ", 0x00000001)
 {
-    /*
-     * iASL Warning: There was 1 external control method found during
-     * disassembly, but only 0 were resolved (1 unresolved). Additional
-     * ACPI tables may be required to properly disassemble the code. This
-     * resulting disassembler output file may not compile because the
-     * disassembler did not know how many arguments to assign to the
-     * unresolved methods. Note: SSDTs can be dynamically loaded at
-     * runtime and may or may not be available via the host OS.
-     *
-     * In addition, the -fe option can be used to specify a file containing
-     * control method external declarations with the associated method
-     * argument counts. Each line of the file must be of the form:
-     *     External (<method pathname>, MethodObj, <argument count>)
-     * Invocation:
-     *     iasl -fe refs.txt -d dsdt.aml
-     *
-     * The following methods were unresolved and many not compile properly
-     * because the disassembler had to guess at the number of arguments
-     * required for each:
-     */
-    External (_SB_.PCI0.S19_.PCNT, MethodObj)    // Warning: Unknown method, guessing 1 arguments
-
     Scope (\)
     {
         OperationRegion (DBG, SystemIO, 0x0402, One)
         Field (DBG, ByteAcc, NoLock, Preserve)
         {
             DBGB,   8
         }

         Method (DBUG, 1, NotSerialized)
         {
             ToHexString (Arg0, Local0)
             ToBuffer (Local0, Local0)
             Local1 = (SizeOf (Local0) - One)
             Local2 = Zero
             While ((Local2 < Local1))
             {
@@ -3322,24 +3300,60 @@
                 Method (DVNT, 2, NotSerialized)
                 {
                     If ((Arg0 & One))
                     {
                         Notify (S00, Arg1)
                     }
                 }

                 Method (PCNT, 0, NotSerialized)
                 {
                     BNUM = One
                     DVNT (PCIU, One)
                     DVNT (PCID, 0x03)
                 }
             }

+            Device (S19)
+            {
+                Name (_ADR, 0x00030001)  // _ADR: Address
+                Name (BSEL, Zero)
+                Device (S00)
+                {
+                    Name (_SUN, Zero)  // _SUN: Slot User Number
+                    Name (_ADR, Zero)  // _ADR: Address
+                    Method (_EJ0, 1, NotSerialized)  // _EJx: Eject Device, x=0-9
+                    {
+                        PCEJ (BSEL, _SUN)
+                    }
+
+                    Method (_DSM, 4, Serialized)  // _DSM: Device-Specific Method
+                    {
+                        Return (PDSM (Arg0, Arg1, Arg2, Arg3, BSEL, _SUN))
+                    }
+                }
+
+                Method (DVNT, 2, NotSerialized)
+                {
+                    If ((Arg0 & One))
+                    {
+                        Notify (S00, Arg1)
+                    }
+                }
+
+                Method (PCNT, 0, NotSerialized)
+                {
+                    BNUM = Zero
+                    DVNT (PCIU, One)
+                    DVNT (PCID, 0x03)
+                }
+            }
+
             Method (PCNT, 0, NotSerialized)
             {
-                ^S19.PCNT (^S10.PCNT ())
+                ^S19.PCNT ()
+                ^S10.PCNT ()
             }
         }
     }
 }

Signed-off-by: Ani Sinha <ani@anisinha.ca>
Message-Id: <20211007135750.1277213-3-ani@anisinha.ca>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Igor Mammedov <imammedo@redhat.com>
2 years agotests/acpi/bios-tables-test: add and allow changes to a new q35 DSDT table blob
Ani Sinha [Thu, 7 Oct 2021 13:57:48 +0000 (19:27 +0530)]
tests/acpi/bios-tables-test: add and allow changes to a new q35 DSDT table blob

We are adding a new unit test to cover the acpi hotplug support in q35 for
multi-function bridges. This test uses a new table DSDT.multi-bridge.
We need to allow changes in DSDT acpi table for addition of this new
unit test.

Signed-off-by: Ani Sinha <ani@anisinha.ca>
Message-Id: <20211007135750.1277213-2-ani@anisinha.ca>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Igor Mammedov <imammedo@redhat.com>
2 years agopci: fix PCI resource reserve capability on BE
Michael S. Tsirkin [Wed, 20 Oct 2021 09:48:54 +0000 (05:48 -0400)]
pci: fix PCI resource reserve capability on BE

PCI resource reserve capability should use LE format as all other PCI
things. If we don't then seabios won't boot:

=== PCI new allocation pass #1 ===
PCI: check devices
PCI: QEMU resource reserve cap: size 10000000000000 type io
PCI: secondary bus 1 size 10000000000000 type io
PCI: secondary bus 1 size 00200000 type mem
PCI: secondary bus 1 size 00200000 type prefmem
=== PCI new allocation pass #2 ===
PCI: out of I/O address space

This became more important since we started reserving IO by default,
previously no one noticed.

Fixes: e2a6290aab ("hw/pcie-root-port: Fix hotplug for PCI devices requiring IO")
Cc: marcel.apfelbaum@gmail.com
Fixes: 226263fb5c ("hw/pci: add QEMU-specific PCI capability to the Generic PCI Express Root Port")
Cc: zuban32s@gmail.com
Fixes: 6755e618d0 ("hw/pci: add PCI resource reserve capability to legacy PCI bridge")
Cc: jing2.liu@linux.intel.com
Tested-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
2 years agovhost-vdpa: multiqueue support
Jason Wang [Wed, 20 Oct 2021 04:56:00 +0000 (12:56 +0800)]
vhost-vdpa: multiqueue support

This patch implements the multiqueue support for vhost-vdpa. This is
done simply by reading the number of queue pairs from the config space
and initialize the datapath and control path net client.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-11-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovirtio-net: vhost control virtqueue support
Jason Wang [Wed, 20 Oct 2021 04:55:59 +0000 (12:55 +0800)]
virtio-net: vhost control virtqueue support

This patch implements the control virtqueue support for vhost. This
requires virtio-net to figure out the datapath queue pairs and control
virtqueue via is_datapath and pass the number of those two types
of virtqueues to vhost_net_start()/vhost_net_stop().

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-10-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovhost: record the last virtqueue index for the virtio device
Jason Wang [Wed, 20 Oct 2021 04:55:58 +0000 (12:55 +0800)]
vhost: record the last virtqueue index for the virtio device

This patch introduces a new field in the vhost_dev structure to record
the last virtqueue index for the virtio device. This will be useful
for the vhost backends with 1:N model to start or stop the device
after all the vhost_dev structures were started or stopped.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-9-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovirtio-net: use "queue_pairs" instead of "queues" when possible
Jason Wang [Wed, 20 Oct 2021 04:55:57 +0000 (12:55 +0800)]
virtio-net: use "queue_pairs" instead of "queues" when possible

Most of the time, "queues" really means queue pairs. So this patch
switch to use "queue_pairs" to avoid confusion.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-8-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovhost-net: control virtqueue support
Jason Wang [Wed, 20 Oct 2021 04:55:56 +0000 (12:55 +0800)]
vhost-net: control virtqueue support

We assume there's no cvq in the past, this is not true when we need
control virtqueue support for vhost-user backends. So this patch
implements the control virtqueue support for vhost-net. As datapath,
the control virtqueue is also required to be coupled with the
NetClientState. The vhost_net_start/stop() are tweaked to accept the
number of datapath queue pairs plus the the number of control
virtqueue for us to start and stop the vhost device.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-7-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agonet: introduce control client
Jason Wang [Wed, 20 Oct 2021 04:55:55 +0000 (12:55 +0800)]
net: introduce control client

This patch introduces a boolean for the device has control queue which
can accepts control command via network queue.

The first user would be the control virtqueue support for vhost.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-6-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovhost-vdpa: let net_vhost_vdpa_init() returns NetClientState *
Jason Wang [Wed, 20 Oct 2021 04:55:54 +0000 (12:55 +0800)]
vhost-vdpa: let net_vhost_vdpa_init() returns NetClientState *

This patch switches to let net_vhost_vdpa_init() to return
NetClientState *. This is used for the callers to allocate multiqueue
NetClientState for multiqueue support.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-5-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovhost-vdpa: prepare for the multiqueue support
Jason Wang [Wed, 20 Oct 2021 04:55:53 +0000 (12:55 +0800)]
vhost-vdpa: prepare for the multiqueue support

Unlike vhost-kernel, vhost-vdpa adapts a single device multiqueue
model. So we need to simply use virtqueue index as the vhost virtqueue
index. This is a must for multiqueue to work for vhost-vdpa.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-4-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovhost-vdpa: classify one time request
Jason Wang [Wed, 20 Oct 2021 04:55:52 +0000 (12:55 +0800)]
vhost-vdpa: classify one time request

Vhost-vdpa uses one device multiqueue queue (pairs) model. So we need
to classify the one time request (e.g SET_OWNER) and make sure those
request were only called once per device.

This is used for multiqueue support.

Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-3-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovhost-vdpa: open device fd in net_init_vhost_vdpa()
Jason Wang [Wed, 20 Oct 2021 04:55:51 +0000 (12:55 +0800)]
vhost-vdpa: open device fd in net_init_vhost_vdpa()

This patch switches to open device fd in net_init_vhost_vpda(). This is
used to prepare for the multiqueue support.

Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211020045600.16082-2-jasowang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agobios-tables-test: don't disassemble empty files
Michael S. Tsirkin [Tue, 19 Oct 2021 10:07:45 +0000 (06:07 -0400)]
bios-tables-test: don't disassemble empty files

A recommended way to populate new tables is to have an
empty expected file. In this case, attempts to disassemble
will fail but it is useful to disassemble the actual files.
Detect and skip decompile step in this case.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agorebuild-expected-aml.sh: allow partial target list
Michael S. Tsirkin [Mon, 18 Oct 2021 21:14:28 +0000 (17:14 -0400)]
rebuild-expected-aml.sh: allow partial target list

Only rebuild AML for configured targets.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agoqdev/qbus: remove failover specific code
Laurent Vivier [Tue, 19 Oct 2021 07:15:32 +0000 (09:15 +0200)]
qdev/qbus: remove failover specific code

Commit f3a850565693 ("qdev/qbus: add hidden device support") has
introduced a generic way to hide a device but it has modified
qdev_device_add() to check a specific option of the failover device,
"failover_pair_id", before calling the generic mechanism.

It's not needed (and not generic) to do that in qdev_device_add() because
this is also checked by the failover_hide_primary_device() function that
uses the generic mechanism to hide the device.

Cc: Jens Freimann <jfreimann@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20211019071532.682717-3-lvivier@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
2 years agovhost-user-blk-test: pass vhost-user socket fds to QSD
Stefan Hajnoczi [Tue, 19 Oct 2021 13:56:55 +0000 (14:56 +0100)]
vhost-user-blk-test: pass vhost-user socket fds to QSD

qemu-storage-daemon is launched with the vhost-user listen socket path.
The path is first unlinked before opening the listen socket. This
prevents stale UNIX domain socket files from stopping socket
initialization.

This behavior is undesirable in vhost-user-blk-test and the cause of a
bug:

There is a race condition in vhost-user-blk-test when QEMU launches
before QSD. It connects to the old socket that QSD unlinks and the
vhost-user connection is never serviced, resulting in a hang.

Pass the listen socket fd to QSD to maintain listen socket continuity
and prevent the lost connection.

Fixes: 806952026df41939680abe92b329715b9b4e01cc ("test: new qTest case to test the vhost-user-blk-server")
Cc: Raphael Norwitz <raphael.norwitz@nutanix.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Thomas Huth <thuth@redhat.com>
Cc: Coiby Xu <coiby.xu@gmail.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20211019135655.83067-1-stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agofailover: fix a regression introduced by JSON'ification of -device
Laurent Vivier [Tue, 19 Oct 2021 07:15:31 +0000 (09:15 +0200)]
failover: fix a regression introduced by JSON'ification of -device

The hide_device helper can be called several times for the same
devices as it shouldn't change any state and should only return an
information.

But not to rely anymore on QemuOpts we have introduced a new field
to store the parameters of the device and don't allow to update it
once it is done.

And as the function is called several times, we ends with:

  warning: Cannot attach more than one primary device to 'virtio0'

That is not only a warning as it prevents to hide the device and breaks
failover.

Fix that by checking the device id.

Now, we fail only if the virtio-net device is really used by two different
devices, for instance:

   -device virtio-net-pci,id=virtio0,failover=on,... \
   -device vfio-pci,id=hostdev0,failover_pair_id=virtio0,... \
   -device e1000e,id=e1000e0,failover_pair_id=virtio0,... \

will exit with:

  Cannot attach more than one primary device to 'virtio0': 'hostdev0' and 'e1000e0'

Fixes: 259a10dbcb4f ("virtio-net: Store failover primary opts pointer locally")
Cc: kwolf@redhat.com
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Message-Id: <20211019071532.682717-2-lvivier@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
2 years agovhost-user: fix duplicated notifier MR init
Xueming Li [Fri, 8 Oct 2021 08:02:15 +0000 (16:02 +0800)]
vhost-user: fix duplicated notifier MR init

In case of device resume after suspend, VQ notifier MR still valid.
Duplicated registrations explode memory block list and slow down device
resume.

Fixes: 44866521bd6e ("vhost-user: support registering external host notifiers")
Cc: tiwei.bie@intel.com
Cc: qemu-stable@nongnu.org
Cc: Yuwei Zhang <zhangyuwei.9149@bytedance.com>
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Message-Id: <20211008080215.590292-1-xuemingl@nvidia.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agodocs: Add documentation for vhost based RNG implementation
Mathieu Poirier [Tue, 12 Oct 2021 20:59:04 +0000 (14:59 -0600)]
docs: Add documentation for vhost based RNG implementation

Add description and example for the vhost-user based RNG implementation.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Message-Id: <20211012205904.4106769-4-mathieu.poirier@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovhost-user-rng-pci: Add vhost-user-rng-pci implementation
Mathieu Poirier [Tue, 12 Oct 2021 20:59:03 +0000 (14:59 -0600)]
vhost-user-rng-pci: Add vhost-user-rng-pci implementation

This patch provides a PCI bus interface to the vhost-user-rng backend.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Message-Id: <20211012205904.4106769-3-mathieu.poirier@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovhost-user-rng: Add vhost-user-rng implementation
Mathieu Poirier [Tue, 12 Oct 2021 20:59:02 +0000 (14:59 -0600)]
vhost-user-rng: Add vhost-user-rng implementation

Introduce a random number generator (RNG) backend that communicates
with a vhost-user server to retrieve entropy.  That way other VMM
that comply with the vhost user protocl can use the same vhost-user
daemon without having to write yet another RNG driver.

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Message-Id: <20211012205904.4106769-2-mathieu.poirier@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agolibvhost-user: fix VHOST_USER_REM_MEM_REG skipping mmap_addr
David Hildenbrand [Mon, 11 Oct 2021 20:10:47 +0000 (22:10 +0200)]
libvhost-user: fix VHOST_USER_REM_MEM_REG skipping mmap_addr

We end up not copying the mmap_addr of all existing regions, resulting
in a SEGFAULT once we actually try to map/access anything within our
memory regions.

Fixes: 875b9fd97b34 ("Support individual region unmap in libvhost-user")
Cc: qemu-stable@nongnu.org
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Raphael Norwitz <raphael.norwitz@nutanix.com>
Cc: "Marc-André Lureau" <marcandre.lureau@redhat.com>
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Coiby Xu <coiby.xu@gmail.com>
Signed-off-by: David Hildenbrand <david@redhat.com>
Message-Id: <20211011201047.62587-1-david@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
2 years agovirtio-iommu: Drop base_name and change generic_name
Eric Auger [Wed, 13 Oct 2021 19:17:55 +0000 (15:17 -0400)]
virtio-iommu: Drop base_name and change generic_name

Drop base_name and turn generic_name into
"virtio-iommu-pci". This is more in line with
other modern-only devices.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Suggested-by: Cornelia Huck <cohuck@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20211013191755.767468-3-eric.auger@redhat.com>
Reviewed-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovirtio-iommu: Remove the non transitional name
Eric Auger [Wed, 13 Oct 2021 19:17:54 +0000 (15:17 -0400)]
virtio-iommu: Remove the non transitional name

Remove the non transitional name for virtio iommu. Like other
devices introduced after 1.0 spec, the virtio-iommu does
not need it.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Reported-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Cornelia Huck <cohuck@redhat.com>
Message-Id: <20211013191755.767468-2-eric.auger@redhat.com>
Reviewed-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2 years agovdpa: Check for iova range at mappings changes
Eugenio Pérez [Thu, 14 Oct 2021 14:12:36 +0000 (16:12 +0200)]
vdpa: Check for iova range at mappings changes

Check vdpa device range before updating memory regions so we don't add
any outside of it, and report the invalid change if any.

Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Message-Id: <20211014141236.923287-4-eperezma@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
2 years agovdpa: Add vhost_vdpa_section_end
Eugenio Pérez [Thu, 14 Oct 2021 14:12:35 +0000 (16:12 +0200)]
vdpa: Add vhost_vdpa_section_end

Abstract this operation, that will be reused when validating the region
against the iova range that the device supports.

Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Message-Id: <20211014141236.923287-3-eperezma@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>