QEMU-Riscv Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Chao Liu <chao.liu@processmission.com>
To: Zephyr Li <fritchleybohrer@gmail.com>
Cc: qemu-devel@nongnu.org, qemu-riscv@nongnu.org,
	qemu-stable@nongnu.org,  pbonzini@redhat.com,
	marcandre.lureau@redhat.com, berrange@redhat.com,
	 pierrick.bouvier@oss.qualcomm.com, alex.bennee@linaro.org,
	palmer@dabbelt.com,  alistair.francis@wdc.com,
	liwei1518@gmail.com, daniel.barboza@oss.qualcomm.com,
	zhiwei_liu@linux.alibaba.com
Subject: Re: [PATCH v2] target/riscv: fix RV32 fixed counter accesses
Date: Fri, 11 Sep 2026 12:44:37 +0800	[thread overview]
Message-ID: <aqOHF2QVtK0UWM-e@MacBook-Pro-4.local> (raw)
In-Reply-To: <20260909072356.42784-1-fritchleybohrer@gmail.com>

On Wed, Sep 09, 2026 at 03:23:55PM +0800, Zephyr Li wrote:
> Since commit cfc96df65e01, riscv_pmu_ctr_get_fixed_counters_val()
> returns the complete 64-bit fixed-counter value. The RV32 counter
> access paths, however, still perform parts of the offset calculation
> on separately extracted 32-bit halves.
> 
> In particular, riscv_pmu_write_ctrh() deposits the low 32 bits of the
> complete fixed-counter value into the high half of mhpmcounter_prev.
> riscv_pmu_read_ctr() also subtracts a 32-bit half of the previous value
> from the complete 64-bit fixed-counter value. Consequently, writes to
> mcycleh can be lost and carries between the low and high halves are not
> handled correctly.
> 
> Keep the fixed-counter offset calculation entirely in 64 bits. Before
> a running counter is partially written, materialize its current
> architectural value and reset the fixed-counter baseline. On reads,
> calculate the complete 64-bit counter value before extracting the half
> requested by RV32.
> 
> Register RV32 system TCG tests with the Meson build. Add a test for
> high-half writes, low-to-high carry, and preserving the carried high
> half across a subsequent low-half write.
> 
> Fixes: cfc96df65e01 ("target/riscv: Remove upper_half from riscv_pmu_ctr_get_fixed_counters_val")
> Resolves: https://gitlab.com/qemu-project/qemu/-/work_items/4219
> Signed-off-by: Zephyr Li <fritchleybohrer@gmail.com>
Reviewed-by: Chao Liu <chao.liu@processmission.com>

Thanks,
Chao

> 
> ---
> Changes in v2:
> - Move the test to a new tests/tcg/riscv32 directory.
> - Convert the test registration to the Meson TCG test framework.
> - Add RV32 TCG cross-compiler options and cover the test directory in
>   MAINTAINERS.
> - Add a riscv32-local semihosting linker script.
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 7183babd6a..5f75b2684f 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -377,6 +377,7 @@ F: include/hw/riscv/
>  F: common-user/host/riscv*
>  F: tests/functional/riscv32
>  F: tests/functional/riscv64
> +F: tests/tcg/riscv32/
>  F: tests/tcg/riscv64/
>  F: tests/qtest/iommu-riscv-test.c
>  
> diff --git a/meson_options.txt b/meson_options.txt
> index 292625af08..2ceebe7a44 100644
> --- a/meson_options.txt
> +++ b/meson_options.txt
> @@ -452,6 +452,10 @@ option('tcg_tests_cross_cc_ppc64le', type: 'string',
>         description: 'cc for ppc64le tcg tests')
>  option('tcg_tests_cross_cflags_ppc64le', type: 'string',
>         description: 'cflags for ppc64le tcg tests')
> +option('tcg_tests_cross_cc_riscv32', type: 'string',
> +       description: 'cc for riscv32 tcg tests')
> +option('tcg_tests_cross_cflags_riscv32', type: 'string',
> +       description: 'cflags for riscv32 tcg tests')
>  option('tcg_tests_cross_cc_riscv64', type: 'string',
>         description: 'cc for riscv64 tcg tests')
>  option('tcg_tests_cross_cflags_riscv64', type: 'string',
> diff --git a/target/riscv/tcg/csr.c b/target/riscv/tcg/csr.c
> index bd4b6dc114..65985efb22 100644
> --- a/target/riscv/tcg/csr.c
> +++ b/target/riscv/tcg/csr.c
> @@ -1337,23 +1337,23 @@ static RISCVException riscv_pmu_write_ctr(CPURISCVState *env, target_ulong val,
>      int deposit_size = rv32 ? 32 : 64;
>      uint64_t ctr;
>  
> -    counter->mhpmcounter_val = deposit64(counter->mhpmcounter_val,
> -                                         0, deposit_size, val);
> -
>      if (!get_field(env->mcountinhibit, BIT(ctr_idx)) &&
>          (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) ||
>           riscv_pmu_ctr_monitor_instructions(env, ctr_idx))) {
>          ctr = riscv_pmu_ctr_get_fixed_counters_val(env, ctr_idx);
> -        counter->mhpmcounter_prev = deposit64(counter->mhpmcounter_prev,
> -                                              0, deposit_size, ctr);
> +        counter->mhpmcounter_val += ctr - counter->mhpmcounter_prev;
> +        counter->mhpmcounter_val = deposit64(counter->mhpmcounter_val,
> +                                             0, deposit_size, val);
> +        counter->mhpmcounter_prev = ctr;
>          if (ctr_idx > 2) {
>              riscv_pmu_setup_timer(env, counter->mhpmcounter_val, ctr_idx);
>          }
>       } else {
> +        counter->mhpmcounter_val = deposit64(counter->mhpmcounter_val,
> +                                             0, deposit_size, val);
>          /* Other counters can keep incrementing from the given value */
>          counter->mhpmcounter_prev = deposit64(counter->mhpmcounter_prev,
>                                                0, deposit_size, val);
> -
>      }
>  
>      return RISCV_EXCP_NONE;
> @@ -1363,20 +1363,22 @@ static RISCVException riscv_pmu_write_ctrh(CPURISCVState *env, target_ulong val,
>                                            uint32_t ctr_idx)
>  {
>      PMUCTRState *counter = &env->pmu_ctrs[ctr_idx];
> -    uint64_t ctrh;
> +    uint64_t ctr;
>  
> -    counter->mhpmcounter_val = deposit64(counter->mhpmcounter_val,
> -                                         32, 32, val);
>      if (!get_field(env->mcountinhibit, BIT(ctr_idx)) &&
>          (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) ||
>           riscv_pmu_ctr_monitor_instructions(env, ctr_idx))) {
> -        ctrh = riscv_pmu_ctr_get_fixed_counters_val(env, ctr_idx);
> -        counter->mhpmcounter_prev = deposit64(counter->mhpmcounter_prev,
> -                                              32, 32, ctrh);
> +        ctr = riscv_pmu_ctr_get_fixed_counters_val(env, ctr_idx);
> +        counter->mhpmcounter_val += ctr - counter->mhpmcounter_prev;
> +        counter->mhpmcounter_val = deposit64(counter->mhpmcounter_val,
> +                                             32, 32, val);
> +        counter->mhpmcounter_prev = ctr;
>          if (ctr_idx > 2) {
>              riscv_pmu_setup_timer(env, counter->mhpmcounter_val, ctr_idx);
>          }
>      } else {
> +        counter->mhpmcounter_val = deposit64(counter->mhpmcounter_val,
> +                                             32, 32, val);
>          counter->mhpmcounter_prev = deposit64(counter->mhpmcounter_prev,
>                                                32, 32, val);
>      }
> @@ -1407,12 +1409,11 @@ RISCVException riscv_pmu_read_ctr(CPURISCVState *env, target_ulong *val,
>      bool rv32 = riscv_cpu_mxl(env) == MXL_RV32;
>      int start = upper_half ? 32 : 0;
>      int length = rv32 ? 32 : 64;
> -    uint64_t ctr_prev, ctr_val;
> +    uint64_t ctr_val;
>  
>      /* Ensure upper_half is only set for XLEN == 32 */
>      g_assert(rv32 || !upper_half);
>  
> -    ctr_prev = extract64(counter->mhpmcounter_prev, start, length);
>      ctr_val  = extract64(counter->mhpmcounter_val, start, length);
>  
>      if (get_field(env->mcountinhibit, BIT(ctr_idx))) {
> @@ -1431,7 +1432,8 @@ RISCVException riscv_pmu_read_ctr(CPURISCVState *env, target_ulong *val,
>      if (riscv_pmu_ctr_monitor_cycles(env, ctr_idx) ||
>          riscv_pmu_ctr_monitor_instructions(env, ctr_idx)) {
>          uint64_t cntr = riscv_pmu_ctr_get_fixed_counters_val(env, ctr_idx) -
> -                                                             ctr_prev + ctr_val;
> +                        counter->mhpmcounter_prev +
> +                        counter->mhpmcounter_val;
>          *val = extract64(cntr, start, length);
>      } else {
>          *val = ctr_val;
> diff --git a/tests/tcg/meson.build b/tests/tcg/meson.build
> index d41a228fb3..60b5ce2529 100644
> --- a/tests/tcg/meson.build
> +++ b/tests/tcg/meson.build
> @@ -169,6 +169,7 @@ subdir('mips64el')
>  subdir('or1k')
>  subdir('ppc64')
>  subdir('ppc64le')
> +subdir('riscv32')
>  subdir('riscv64')
>  subdir('s390x')
>  subdir('sh4')
> diff --git a/tests/tcg/riscv32/meson.build b/tests/tcg/riscv32/meson.build
> new file mode 100644
> index 0000000000..c08dfd772d
> --- /dev/null
> +++ b/tests/tcg/riscv32/meson.build
> @@ -0,0 +1,7 @@
> +# SPDX-License-Identifier: GPL-2.0-or-later
> +
> +cc = 'riscv64-linux-gnu-gcc'
> +cc_dockerfile = 'debian-all-test-cross'
> +cc_docker_host_arch = ['aarch64', 'x86_64']
> +
> +subdir('system')
> diff --git a/tests/tcg/riscv32/semihost.ld b/tests/tcg/riscv32/semihost.ld
> new file mode 100644
> index 0000000000..874838a865
> --- /dev/null
> +++ b/tests/tcg/riscv32/semihost.ld
> @@ -0,0 +1,23 @@
> +/* SPDX-License-Identifier: GPL-2.0-or-later */
> +
> +ENTRY(_start)
> +
> +SECTIONS
> +{
> +    /* virt machine, RAM starts at 2gb */
> +    . = 0x80000000;
> +    .text : {
> +        *(.text)
> +    }
> +    .rodata : {
> +        *(.rodata)
> +    }
> +    /* align r/w section to next 2mb */
> +    . = ALIGN(1 << 21);
> +    .data : {
> +        *(.data)
> +    }
> +    .bss : {
> +        *(.bss)
> +    }
> +}
> diff --git a/tests/tcg/riscv32/system/meson.build b/tests/tcg/riscv32/system/meson.build
> new file mode 100644
> index 0000000000..16f9a06c94
> --- /dev/null
> +++ b/tests/tcg/riscv32/system/meson.build
> @@ -0,0 +1,37 @@
> +# SPDX-License-Identifier: GPL-2.0-or-later
> +
> +tests = []
> +
> +link_script = files('../semihost.ld')[0]
> +cflags = ['-march=rv32im_zicsr',
> +          '-mabi=ilp32',
> +          '-nostdlib',
> +          '-ffreestanding',
> +          '-Wa,--noexecstack',
> +          '-Wl,-T', link_script]
> +qemu_args = ['-M', 'virt',
> +             '-display', 'none',
> +             '-serial', 'stdio',
> +             '-semihosting',
> +             '-bios']
> +
> +tests += {
> +  'test-mcycle.S': {
> +    'cflags': cflags,
> +    'qemu_args': ['-icount', 'shift=1', qemu_args],
> +  },
> +}
> +
> +if 'qemu-system-riscv32' in emulators
> +  tcg_tests += {
> +    'riscv32-softmmu': {
> +      'cc': cc,
> +      'cc_dockerfile': cc_dockerfile,
> +      'cc_docker_host_arch': cc_docker_host_arch,
> +      'folder': 'riscv32',
> +      'gdb_arch': 'riscv32',
> +      'qemu': emulators['qemu-system-riscv32'],
> +      'tests': tests,
> +    }
> +  }
> +endif
> diff --git a/tests/tcg/riscv32/test-mcycle.S b/tests/tcg/riscv32/test-mcycle.S
> new file mode 100644
> index 0000000000..189d1e13a2
> --- /dev/null
> +++ b/tests/tcg/riscv32/test-mcycle.S
> @@ -0,0 +1,45 @@
> +/* SPDX-License-Identifier: GPL-2.0-or-later */
> +
> +	.option	norvc
> +
> +	.text
> +	.global	_start
> +_start:
> +	/* Exercise writes while mcycle is running. */
> +	csrw	mcountinhibit, zero
> +	csrw	mcycle, zero
> +
> +	/* A write to the high half must be immediately observable. */
> +	li	s0, 0x1234ffff
> +	csrw	mcycleh, s0
> +	csrr	t0, mcycleh
> +	bne	t0, s0, fail
> +
> +	/* Check carry from the low half into the high half. */
> +	li	s0, 0x12345678
> +	csrw	mcycleh, s0
> +	li	t0, 0xfffffff0
> +	csrw	mcycle, t0
> +	.rept	32
> +	nop
> +	.endr
> +	csrr	t0, mcycleh
> +	addi	s0, s0, 1
> +	bne	t0, s0, fail
> +
> +	/* A low-half write must preserve the carried high half. */
> +	li	t0, 0x22222222
> +	csrw	mcycle, t0
> +	csrr	t0, mcycleh
> +	bne	t0, s0, fail
> +
> +	li	t0, 0x100000
> +	li	t1, 0x5555	/* FINISHER_PASS */
> +	sw	t1, 0(t0)
> +	j	.
> +
> +fail:
> +	li	t0, 0x100000
> +	li	t1, 0x13333	/* status = FINISHER_FAIL, code = 1 */
> +	sw	t1, 0(t0)
> +	j	.
> -- 
> 2.43.0


      parent reply	other threads:[~2026-09-11  4:45 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-09  7:23 [PATCH v2] target/riscv: fix RV32 fixed counter accesses Zephyr Li
2026-09-09 19:58 ` Daniel Henrique Barboza
2026-09-11  4:44 ` Chao Liu [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=aqOHF2QVtK0UWM-e@MacBook-Pro-4.local \
    --to=chao.liu@processmission.com \
    --cc=alex.bennee@linaro.org \
    --cc=alistair.francis@wdc.com \
    --cc=berrange@redhat.com \
    --cc=daniel.barboza@oss.qualcomm.com \
    --cc=fritchleybohrer@gmail.com \
    --cc=liwei1518@gmail.com \
    --cc=marcandre.lureau@redhat.com \
    --cc=palmer@dabbelt.com \
    --cc=pbonzini@redhat.com \
    --cc=pierrick.bouvier@oss.qualcomm.com \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-riscv@nongnu.org \
    --cc=qemu-stable@nongnu.org \
    --cc=zhiwei_liu@linux.alibaba.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox