From: Namhyung Kim <namhyung@kernel.org>
To: Tengda Wu <wutengda@huaweicloud.com>
Cc: james.clark@linaro.org, xueshuai@linux.alibaba.com,
Li Huafei <lihuafei1@huawei.com>,
Peter Zijlstra <peterz@infradead.org>,
leo.yan@linux.dev, Ian Rogers <irogers@google.com>,
Kim Phillips <kim.phillips@arm.com>,
Mark Rutland <mark.rutland@arm.com>,
Arnaldo Carvalho de Melo <acme@kernel.org>,
Ingo Molnar <mingo@redhat.com>, Bill Wendling <morbo@google.com>,
Nick Desaulniers <nick.desaulniers+lkml@gmail.com>,
Alexander Shishkin <alexander.shishkin@linux.intel.com>,
Adrian Hunter <adrian.hunter@intel.com>,
Zecheng Li <zli94@ncsu.edu>,
linux-perf-users@vger.kernel.org, linux-kernel@vger.kernel.org,
llvm@lists.linux.dev
Subject: Re: [PATCH v3 20/21] perf annotate-arm64: Support per-cpu variable access tracking
Date: Thu, 9 Jul 2026 00:29:20 -0700 [thread overview]
Message-ID: <ak9N0CY1sAtGt47I@z2> (raw)
In-Reply-To: <20260701035355.752944-21-wutengda@huaweicloud.com>
On Wed, Jul 01, 2026 at 03:53:54AM +0000, Tengda Wu wrote:
> Extend update_insn_state() for arm64 to handle per-cpu variable
> addressing.
>
> On arm64, per-cpu variables are accessed by adding a per-cpu offset
> (typically from the '__per_cpu_offset' array) to the address of a global
> variable or a local variable with '__percpu' modifier(e.g.,
> 's32 __percpu *counters' in struct percpu_counter). This results in
> instruction patterns like:
>
> ldr x0, [x6, x5] // Pattern A: direct load per-cpu instance
> add x0, x6, x5 // Pattern B: compute per-cpu address
>
> where x6 holds the per-cpu offset retrieved from:
>
> adrp x4, <page>
> add x4, x4, #offset // x4 = &__per_cpu_offset
> ldr x6, [x4, w0, sxtw #3] // x6 = __per_cpu_offset[cpu]
>
> and x5 is one of the following:
>
> case 1: global variable
> adrp x5, <page>
> add x5, x5, #offset // x5 = &global_var
>
> case 2: local variable with '__percpu' modifier
> ldr x1, [x25, #32] // x1 = local_percpu_ptr
>
> To handle such cases:
>
> 1. Identify per-cpu base initialization: Detect 'adrp+add' pairs that
> resolve to the '__per_cpu_offset' symbol and mark the destination
> register as TSR_KIND_PERCPU_BASE.
> 2. Propagate type information: During subsequent 'ldr' or 'add'
> instructions, if one register is TSR_KIND_PERCPU_BASE, attempt to
> resolve the type from the other register.
>
> A real-world example is shown below:
>
> ffff8000808f2d28 <cppc_set_perf>:
> ffff8000808f2d38: adrp x2, ffff800082033000
> ffff8000808f2d3c: add x5, x2, #0x3f8 // x5 = &__per_cpu_offset
> ffff8000808f2d44: adrp x2, ffff800081f73000
> ffff8000808f2d48: add x2, x2, #0x6b8 // x2 = &cpu_pcc_subspace_idx
> ffff8000808f2d6c: ldr x5, [x5, w0, sxtw #3] // x5 = __per_cpu_offset[cpu]
> * ffff8000808f2d80: ldr w23, [x5, x2] // per_cpu(cpu_pcc_subspace_idx, cpu)
>
> Before this commit, the tracker could not link x5 back to a per-cpu
> context, resulting in an incorrect data type resolution:
>
> adrp [10] global addr=0xffff800082033000 -> reg2
> add [14] global 0x3f8(reg2) -> reg5
> adrp [1c] global addr=0xffff800081f73000 -> reg2
> add [20] global 0x6b8(reg2) -> reg2
> ldr [44] global (reg5, reg0) -> reg5 type='long unsigned int[]' size=0x1000
> chk [58] reg5 offset=0 ok=1 kind=1 (long unsigned int[]) : Good!
> found by insn track: 0(reg5, reg2) type-offset=0
> final result: type='long unsigned int' size=0x8
>
> After this commit, the tracker correctly identifies the per-cpu flow and
> resolves the actual variable type:
>
> ldr [44] global (reg5, reg0) -> reg5 percpu base
> chk [58] reg5 offset=0 ok=1 kind=2 percpu var : retry
> chk [58] reg2 offset=0 ok=1 kind=7 global addr : Good!
> found by insn track: 0(reg5, reg2) type-offset=0
> final result: type='int' size=0x4
>
> Signed-off-by: Tengda Wu <wutengda@huaweicloud.com>
> ---
> .../perf/util/annotate-arch/annotate-arm64.c | 60 ++++++++++++++++++-
> tools/perf/util/annotate-data.c | 24 +++++++-
> 2 files changed, 82 insertions(+), 2 deletions(-)
>
> diff --git a/tools/perf/util/annotate-arch/annotate-arm64.c b/tools/perf/util/annotate-arch/annotate-arm64.c
> index 6f96e75d313d..ec6fd59d51a2 100644
> --- a/tools/perf/util/annotate-arch/annotate-arm64.c
> +++ b/tools/perf/util/annotate-arch/annotate-arm64.c
> @@ -462,6 +462,15 @@ static void update_load_insn_state(struct type_state *state,
> u64 ip = dloc->ms->sym->start + dl->al.offset;
> u64 addr = src_tsr.addr + reg_offset;
> int offset;
> + u8 kind;
> + const char *var_name = NULL;
> +
> + /* it might be per-cpu offset */
> + if (get_global_var_info(dloc, addr, &var_name, &offset) &&
> + !strcmp(var_name, "__per_cpu_offset"))
> + kind = TSR_KIND_PERCPU_BASE;
It seems you need to check if the target DSO is a kernel or a module by
calling `dso__kernel(map__dso(dloc->ms->map))` first.
Thanks,
Namhyung
> + else
> + kind = TSR_KIND_TYPE;
>
> if (!get_global_var_type(cu_die, dloc, ip, addr, &offset, &type_die) ||
> !die_get_member_type(&type_die, offset, &type_die)) {
> @@ -470,7 +479,7 @@ static void update_load_insn_state(struct type_state *state,
> }
>
> tsr->type = type_die;
> - tsr->kind = TSR_KIND_TYPE;
> + tsr->kind = kind;
> tsr->offset = 0;
> tsr->addr = 0;
> tsr->ok = true;
> @@ -484,6 +493,28 @@ static void update_load_insn_state(struct type_state *state,
> }
> pr_debug_type_name(&tsr->type, tsr->kind);
> }
> + /* Or check if it's a per-cpu access */
> + else if (src_tsr.kind == TSR_KIND_PERCPU_BASE) {
> + int reg2;
> +
> + if (!src->multi_regs || src->reg1 == src->reg2 ||
> + sreg == src->reg2 /* retried */) {
> + invalidate_reg_state(tsr);
> + goto out_adjust;
> + }
> +
> + reg2 = src->reg2;
> + if (!has_reg_type(state, reg2) || !state->regs[reg2].ok ||
> + (state->regs[reg2].kind != TSR_KIND_GLOBAL_ADDR &&
> + state->regs[reg2].kind != TSR_KIND_TYPE)) {
> + invalidate_reg_state(tsr);
> + goto out_adjust;
> + } else {
> + /* Treat percpu as array: resolve type from reg2 */
> + sreg = src->reg2;
> + goto retry;
> + }
> + }
> /* Or try another register if any */
> else if (src->multi_regs && src->reg1 != src->reg2 && sreg != src->reg2) {
> sreg = src->reg2;
> @@ -693,6 +724,33 @@ static void update_add_insn_state(struct type_state *state,
> pr_debug_dtp("add [%x] global %#x(reg%d) -> reg%d\n",
> insn_offset, reg_offset, sreg, dreg);
> }
> + /* Handle per-cpu base address calculation for per-cpu variables */
> + else if (src_tsr.kind == TSR_KIND_PERCPU_BASE) {
> + /*
> + * A per-cpu base acts like an array index. Adding it to
> + * a global variable or typed register should preserve
> + * the original variable's type. Inherit the type from
> + * the typed register.
> + */
> + if (!src->multi_regs || !has_reg_type(state, src->reg2) ||
> + !state->regs[src->reg2].ok ||
> + (state->regs[src->reg2].kind != TSR_KIND_GLOBAL_ADDR &&
> + state->regs[src->reg2].kind != TSR_KIND_TYPE)) {
> + invalidate_reg_state(tsr);
> + return;
> + }
> +
> + tsr->type = state->regs[src->reg2].type;
> + tsr->kind = state->regs[src->reg2].kind;
> + tsr->offset = state->regs[src->reg2].offset;
> + tsr->addr = state->regs[src->reg2].addr;
> + tsr->ok = state->regs[src->reg2].ok;
> +
> + pr_debug_dtp("add [%x] percpu (reg%d) -> reg%d",
> + insn_offset, sreg, dreg);
> + pr_debug_type_name(&tsr->type, tsr->kind);
> + return;
> + }
> /* Or try another register if any */
> else if (src->multi_regs && src->reg1 != src->reg2 && sreg != src->reg2) {
> sreg = src->reg2;
> diff --git a/tools/perf/util/annotate-data.c b/tools/perf/util/annotate-data.c
> index 6fa5cd373a46..b0bf0437168b 100644
> --- a/tools/perf/util/annotate-data.c
> +++ b/tools/perf/util/annotate-data.c
> @@ -1252,11 +1252,33 @@ static enum type_match_result check_matching_type(struct type_state *state,
> }
>
> if (state->regs[reg].kind == TSR_KIND_PERCPU_BASE) {
> - u64 var_addr = dloc->op->offset;
> + u64 var_addr;
> int var_offset;
>
> pr_debug_dtp("percpu var");
>
> + if (arch__is_arm64(dloc->arch)) {
> + int reg2;
> +
> + if (!dloc->op->multi_regs ||
> + dloc->op->reg1 == dloc->op->reg2 || !retry)
> + return PERF_TMR_BAIL_OUT;
> +
> + reg2 = dloc->op->reg2;
> + if (!has_reg_type(state, reg2) ||
> + !state->regs[reg2].ok ||
> + (state->regs[reg2].kind != TSR_KIND_GLOBAL_ADDR &&
> + state->regs[reg2].kind != TSR_KIND_TYPE))
> + return PERF_TMR_NO_TYPE;
> +
> + pr_debug_dtp(" : retry\n");
> + retry = false;
> + reg = reg2;
> + goto again;
> + }
> +
> + var_addr = dloc->op->offset;
> +
> if (dloc->op->multi_regs) {
> int reg2 = dloc->op->reg2;
>
> --
> 2.34.1
>
next prev parent reply other threads:[~2026-07-09 7:29 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-01 3:53 [PATCH v3 00/21] perf arm64: Support data type profiling Tengda Wu
2026-07-01 3:53 ` [PATCH v3 01/21] perf capstone: Fix kernel map reference count leak Tengda Wu
2026-07-09 5:57 ` Namhyung Kim
2026-07-10 21:49 ` Namhyung Kim
2026-07-01 3:53 ` [PATCH v3 02/21] perf capstone: Fix arm64 jump/adrp disassembly mismatch with objdump Tengda Wu
2026-07-01 3:53 ` [PATCH v3 03/21] perf llvm: Fix arm64 adrp instruction " Tengda Wu
2026-07-09 6:18 ` Namhyung Kim
2026-07-09 7:49 ` Tengda Wu
2026-07-01 3:53 ` [PATCH v3 04/21] perf annotate-arm64: Generalize arm64_mov__parse to support more instructions Tengda Wu
2026-07-01 3:53 ` [PATCH v3 05/21] perf annotate-arm64: Handle load and store instructions Tengda Wu
2026-07-01 3:53 ` [PATCH v3 06/21] perf dwarf-regs: Adapt get_dwarf_regnum() for arm64 Tengda Wu
2026-07-01 3:53 ` [PATCH v3 07/21] perf annotate: Adapt arch__dwarf_regnum() " Tengda Wu
2026-07-01 3:53 ` [PATCH v3 08/21] perf annotate: Introduce extract_op_location callback for arch-specific parsing Tengda Wu
2026-07-01 3:53 ` [PATCH v3 09/21] perf annotate-arm64: Implement extract_op_location() callback Tengda Wu
2026-07-01 3:53 ` [PATCH v3 10/21] perf annotate: Deduplicate overlapping ARM SPE events for data type profiling Tengda Wu
2026-07-01 3:53 ` [PATCH v3 11/21] perf auxtrace: Set default period to 1 for PERF_ITRACE_PERIOD_INSTRUCTIONS type Tengda Wu
2026-07-01 3:53 ` [PATCH v3 12/21] perf annotate-data: Extract invalidate_reg_state() as a common helper Tengda Wu
2026-07-01 3:53 ` [PATCH v3 13/21] perf annotate-arm64: Enable instruction tracking support Tengda Wu
2026-07-01 3:53 ` [PATCH v3 14/21] perf annotate-arm64: Support load instruction tracking Tengda Wu
2026-07-01 3:53 ` [PATCH v3 15/21] perf annotate-arm64: Support store " Tengda Wu
2026-07-01 3:53 ` [PATCH v3 16/21] perf annotate-arm64: Support stack variable tracking Tengda Wu
2026-07-01 3:53 ` [PATCH v3 17/21] perf annotate-arm64: Support 'mov' instruction tracking Tengda Wu
2026-07-01 3:53 ` [PATCH v3 18/21] perf annotate-arm64: Support 'add' " Tengda Wu
2026-07-01 3:53 ` [PATCH v3 19/21] perf annotate-arm64: Support 'adrp' instruction to track global variables Tengda Wu
2026-07-09 7:31 ` Namhyung Kim
2026-07-09 7:42 ` Tengda Wu
2026-07-01 3:53 ` [PATCH v3 20/21] perf annotate-arm64: Support per-cpu variable access tracking Tengda Wu
2026-07-09 7:29 ` Namhyung Kim [this message]
2026-07-01 3:53 ` [PATCH v3 21/21] perf annotate-arm64: Support 'mrs' instruction to track 'current' pointer Tengda Wu
2026-07-09 5:54 ` [PATCH v3 00/21] perf arm64: Support data type profiling Namhyung Kim
2026-07-09 8:01 ` Tengda Wu
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=ak9N0CY1sAtGt47I@z2 \
--to=namhyung@kernel.org \
--cc=acme@kernel.org \
--cc=adrian.hunter@intel.com \
--cc=alexander.shishkin@linux.intel.com \
--cc=irogers@google.com \
--cc=james.clark@linaro.org \
--cc=kim.phillips@arm.com \
--cc=leo.yan@linux.dev \
--cc=lihuafei1@huawei.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-perf-users@vger.kernel.org \
--cc=llvm@lists.linux.dev \
--cc=mark.rutland@arm.com \
--cc=mingo@redhat.com \
--cc=morbo@google.com \
--cc=nick.desaulniers+lkml@gmail.com \
--cc=peterz@infradead.org \
--cc=wutengda@huaweicloud.com \
--cc=xueshuai@linux.alibaba.com \
--cc=zli94@ncsu.edu \
/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