From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.remote-ip=74.125.82.175 ARC-Seal: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; cv=none; t=1794850732; b=Y9LGXCFDEWCPxx65bugI7JHtOsz5Bqx4ZLMg+bBpGiWfHD0X9Wcn0M4xZ1dTrQ5jGx33 Ck1JkMX8ppllVowp01FdoY1+kbYsf/wGok+j8jeT5QvwyvbLsFXBXHaIf31atp6ElXdAY 0cvIPvOlU9MfGuhDtfypd44aBqpfx93RqNNbeic45p4AA6zxfZJZpES5p2zuAP6UTxXO1 vooV+HtTLlqvk7+uspnqQPsA9w+fLZXf0MnQ5cwXdrquvQTVp4BsWh6sNIw0ipbYLZff+ ed+CbCgvYMRn+CFWAKxIR0Sfxd4toFoAY6Lf6iPkalgbm/ArxfrcuyfnG0qvC4dJWJg== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1794850732; h=DMARC-Filter:Received:DKIM-Signature:X-Google-DKIM-Signature: X-Gm-Message-State:X-Gm-Gg:X-Received:Received:From:To:Cc:Subject: Date:Message-ID:X-Mailer:In-Reply-To:References:MIME-Version: Content-Transfer-Encoding; bh=SzqhHzIyUkPcykE/zhmvkeMuT8irPjp2F6ctpeBwX0E=; b=oZRpf6Irbn4H9f1Ce2dS2A4qSOZTNZ5ajNjC3ldGAWsH9kq/CsZkKGRzpruMz95IRc56 wLsYtmozUBPp7itCwWeSmERbycaRMhx1HWyQ2tiMomJzbmuSFukju21Lpv0gwsAd6AMIP q3m1xcGhLwBzUoqf4z/UlRe0nj7FmZWc3x59P76UjmnXuYES3gTkZuKAxQnsf3ls4IXLQ pQJJI34TYY03/1RsmqB87bmot+vyHAFFRrAXUGwrsHG83nKq6TAoOwmn1pG3wHHt/Z6JU cHIAtjp4z2fA2bouAFYd7GtLqKeoT/b5MevoQGtG6tAeZTPretm494Lz0LyuwL+CLdg== ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass header.from=gmail.com; spf=pass smtp.mailfrom=gmail.com; dkim=pass header.d=gmail.com header.i=@gmail.com header.a=rsa-sha256 header.s=20260707 header.b=mkDP4afW; arc=none smtp.remote-ip=74.125.82.175 Received: from mail-dy1-f175.google.com (mail-dy1-f175.google.com [74.125.82.175]) (using TLSv1.3 with cipher TLS_AES_128_GCM_SHA256 (128/128 bits) key-exchange X25519 server-signature ECDSA (prime256v1) server-digest SHA256) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id D3362606A332 for ; Mon, 16 Nov 2026 17:38:51 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.4.2 smtp.subspace.kernel.org D3362606A332 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=gmail.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.a=rsa-sha256 header.s=20260707 header.b=mkDP4afW Received: by mail-dy1-f175.google.com with SMTP id 5a478bee46e88-3712b6d9843so3039162eec.1 for ; Mon, 16 Nov 2026 09:38:51 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20260707; t=1794850731; x=1795455531; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to:content-type; bh=SzqhHzIyUkPcykE/zhmvkeMuT8irPjp2F6ctpeBwX0E=; b=mkDP4afWDnm6Q8euc90fQxPyW3tf5NJCSGfzpquBMBvdl/KSNeOpat8X6UcltgzKdh jhy0FFqCSDkZSnpR3geXQaJOh0mihLC4jB7JWkbyk/X04AGcd/tp71Etv7m0wpfysxbv tPhFUKaN1OVlvWggzm8AUMvcbZC/TNKqbMSSKXSN9AtsHcttgkKLIe4hIgdxSAlZCWU0 mRHSWx342KenTfsK4kr+jWaOPUxDCTOfRqxMhQbEx1BU2q7PWTzFAz7BSQjy1uPtloIl tdySxAid+j5Tb+GYO1uqA6fomqfVDKuOPpKu3EZUKIJq2SYbm4z0dvDNu8L4fWmCKwBL 7JLw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20260707; t=1794850731; x=1795455531; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-gg:x-gm-message-state:from :to:cc:subject:date:message-id:reply-to:content-type; bh=SzqhHzIyUkPcykE/zhmvkeMuT8irPjp2F6ctpeBwX0E=; b=rOAktUGXNQZhZhwlE3WES7N0DiTYsfujGMxtqweBJpMcsjuEGcC/76rrps5QDa989U awSHLbMLX7n/V3w7cQUHi05AnvpFJkcm7F+yNnXSF5T++BP4aieKUfVlD0Jjtw+pa3ap oqFgBmwt8UU2nsU0lkJJIghKitD5eLqmBaP/peCer7qyWRZphl3C4Zn8OZy81tiwQu1Z eKcxKholedXfjndDyOieUAuJOL9T4xOOU3iJ4KuVntYfsoRAqJZVchKETfuRTKfogPSA GB8Gfv1njFVAmJ6uNRBfjYOIyNv2Swt+u7Qh6U3x3QaFCjjAzoiKRA9cfalGjFdFAT1w dEzw== X-Gm-Message-State: APuU4Z9fOem1s87ZAbPTkrfyCgQCeQIWQhjvpj70Ut6Y1sry/QssYsLd kaeWVsQ5MsAD/OY22eVqFos6z7cqwV4AmNrh53ERCC8N9vGGHbMvDc7vtKZXcw== X-Gm-Gg: AYBFou22ESZL/tagz4sVjEoL7zjFHF2Ng/Xng74+SMLCEX414zPN509s+1it1U/Uxo7 nE5tR1ThO+u8jGFXZw9miFov94FW0+lsvGySWqcGiHmlPOIb09S2iaZHXB+na3G9kmYX6R1gCGh GtR72UJQLxYgrR54qc42Idq7u7HvDDoBlrD5X60oypLW/Xwyg/3PTS/9TwJSE20JBK/c2iSY+B1 mxWbfGS/p1Wty3ajiEiKB6IMsnEkESMCjIbbTEgdhdtxqtm1qluYIlw4GkZ2rjpO00BBwTZPNMY porq5ut3/jRj9qjScySUYQ0kk5X2OojDiFgllmaOS6ZqsttmzIGPxmP8ULr5MtanCNucc6+Lje/ CsovTAek61/9Lq8uUdzYQy2JssAqXiMWQGq8jjrIzcudFit+mdqlFMnVJ3JmykY9TtHZqfdk3n2 AfggSeoUHhtcdLcP4KqGzjUBC6gvEq2R7zPqmahqB0dSJUfJqpzJDl5rHIwFRyVJ+Dj32BM5jwo /Bqnjr5ify1QawlfbjiccAzRHKmyiF2EZf3idNoDLs2kwZvnEnT2jbr6sNxrbdZVoUlnvjaLQvi 2xl3nMhTXghooE2bisz6gps2J6t0MA== X-Received: by 2002:a05:693c:8852:b0:36a:1452:8dad with SMTP id 5a478bee46e88-37966f32039mr11206896eec.20.1794850731028; Mon, 16 Nov 2026 09:38:51 -0800 (PST) Received: from eldorado ([2600:8802:b00:1f88::287]) by smtp.gmail.com with ESMTPSA id 5a478bee46e88-37955e0d966sm36635215eec.15.2026.11.16.09.38.49 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 16 Nov 2026 09:38:50 -0800 (PST) From: Florian Fainelli To: linux-mips@vger.kernel.org Cc: Florian Fainelli , Thomas Bogendoerfer , Jonathan Corbet , Shuah Khan , Randy Dunlap , Peter Zijlstra , Ingo Molnar , Arnaldo Carvalho de Melo , Namhyung Kim , Mark Rutland , Alexander Shishkin , Jiri Olsa , Ian Rogers , Adrian Hunter , James Clark , Thomas Gleixner , linux-doc@vger.kernel.org (open list:DOCUMENTATION), linux-kernel@vger.kernel.org (open list), linux-perf-users@vger.kernel.org (open list:PERFORMANCE EVENTS SUBSYSTEM), "Maciej W. Rozycki" Subject: [PATCH v3 2/3] MIPS: perf: Add SB1 userspace callchain unwinding Date: Mon, 16 Nov 2026 09:38:37 -0800 Message-ID: <20261116173843.911523-3-f.fainelli@gmail.com> X-Mailer: git-send-email 2.53.0 In-Reply-To: <20261116173843.911523-1-f.fainelli@gmail.com> References: <20261116173843.911523-1-f.fainelli@gmail.com> X-Mailing-List: linux-mips@vger.kernel.org List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Restore the userspace callchain capability formerly available through MIPS OProfile, exposing it through perf for SB1 and SB1A. Commit 5f307491f3a0 ("MIPS: oprofile: Add callgraph support") introduced an in-kernel userspace prologue scanner, which disappeared with the removal of MIPS OProfile support in e258958945c6 ("arch: mips: Remove CONFIG_OPROFILE support"). The MIPS perf userspace callchain hook remained unimplemented. Follow the former OProfile approach of examining instructions before the sampled PC to find standard MIPS stack adjustments and saved return addresses. Bound each scan to 128 instructions, use nofault user reads, validate frame data and stack progress, and stop at the perf stack limit. Handle o32, n32 and n64 address widths and the actual sw/sd return-address slot widths, rather than relying on the kernel's unsigned long size. Allow the sampled leaf to use the live return-address register, and permit repeated return PCs when the stack advances so recursion is not truncated. This complements the existing MIPS DWARF callchain mechanism, enabled by 1ddc96bd42da ("MIPS: kernel: Support extracting off-line stack traces from user-space with perf") and b5f184fbdb03 ("perf tools: Support MIPS unwinding and dwarf-regs"). With --call-graph dwarf, the kernel records user registers and a stack snapshot for the perf tool to unwind using DWARF information. The new walker instead produces PERF_SAMPLE_CALLCHAIN entries in the kernel for ordinary perf -g/--call-graph fp sampling, without requiring DWARF unwind information or a userspace unwinding library. It does not change or replace the DWARF path, which can handle frames beyond the standard prologues recognized by this best-effort scanner. Limit enablement to SB1 and SB1A, whose standard instruction encoding is handled here and which have neither MIPS16 nor microMIPS. Other MIPS CPU types retain their existing userspace callchain behavior. Nonstandard prologues, dynamic stack adjustments and signal trampolines can truncate chains. Signed-off-by: Florian Fainelli --- arch/mips/kernel/perf_event.c | 110 ++++++++++++++++++++++++++++++++-- 1 file changed, 106 insertions(+), 4 deletions(-) diff --git a/arch/mips/kernel/perf_event.c b/arch/mips/kernel/perf_event.c index 5d7a9c03903b..cb2099416ac6 100644 --- a/arch/mips/kernel/perf_event.c +++ b/arch/mips/kernel/perf_event.c @@ -13,15 +13,117 @@ #include #include +#include +#include +#include #include /* Callchain handling code. */ -/* - * Leave userspace callchain empty for now. When we find a way to trace - * the user stack callchains, we will add it here. - */ +struct sb1_user_frame { + unsigned long pc; + unsigned long sp; + unsigned long ra; +}; + +static unsigned long sb1_user_address(unsigned long addr) +{ + if (IS_ENABLED(CONFIG_64BIT) && test_thread_flag(TIF_32BIT_ADDR)) + return (u32)addr; + return addr; +} + +/* Best-effort prologue unwinding, as provided by the former MIPS OProfile. */ +static bool sb1_unwind_user(struct sb1_user_frame *frame) +{ + unsigned long pc = frame->pc, caller = frame->ra; + unsigned int frame_size = 0, ra_size = 0, ra_offset = 0, n; + union mips_instruction insn; + + /* SB1 has neither MIPS16 nor microMIPS instructions. */ + if ((pc & 3) || (frame->sp & 3)) + return false; + + for (n = 0; n < 128 && pc >= sizeof(insn); n++) { + pc -= sizeof(insn); + if (copy_from_user_nofault(&insn, (void __user *)pc, sizeof(insn))) + break; + if (insn.i_format.rs == 29 && insn.i_format.rt == 29 && + (insn.i_format.opcode == addiu_op || + insn.i_format.opcode == daddiu_op)) { + /* A positive adjustment belongs to an epilogue. */ + if (insn.i_format.simmediate >= 0) + break; + frame_size = -insn.i_format.simmediate; + break; + } + if (insn.i_format.rs == 29 && insn.i_format.rt == 31 && + insn.i_format.simmediate >= 0) { + if (insn.i_format.opcode == sw_op) + ra_size = sizeof(u32); + else if (insn.i_format.opcode == sd_op) + ra_size = sizeof(u64); + else + continue; + ra_offset = insn.i_format.simmediate; + } + /* Do not scan into the preceding function through a return. */ + if (insn.r_format.opcode == spec_op && + insn.r_format.func == jr_op) + break; + } + + if (ra_size) { + u64 ra64; + u32 ra32; + void __user *slot = (void __user *)(frame->sp + ra_offset); + + if (!frame_size || ra_offset + ra_size > frame_size) + return false; + if (ra_size == sizeof(ra32)) { + if (copy_from_user_nofault(&ra32, slot, sizeof(ra32))) + return false; + caller = ra32; + } else { + if (copy_from_user_nofault(&ra64, slot, sizeof(ra64))) + return false; + caller = ra64; + } + } + + caller = sb1_user_address(caller); + if (!caller || (caller == frame->pc && !frame_size) || (caller & 3) || + !access_ok((void __user *)caller, sizeof(insn)) || + frame->sp + frame_size < frame->sp) + return false; + frame->sp += frame_size; + frame->pc = caller; + /* Only the sampled leaf may use the live return-address register. */ + frame->ra = 0; + return true; +} + +void perf_callchain_user(struct perf_callchain_entry_ctx *entry, + struct pt_regs *regs) +{ + struct sb1_user_frame frame = { + .pc = sb1_user_address(regs->cp0_epc), + .sp = sb1_user_address(regs->regs[29]), + .ra = sb1_user_address(regs->regs[31]), + }; + unsigned long saved_baduaddr; + + if (current_cpu_type() != CPU_SB1 && current_cpu_type() != CPU_SB1A) + return; + /* Nofault reads must not clobber an interrupted usercopy's fault address. */ + saved_baduaddr = current->thread.cp0_baduaddr; + do { + if (perf_callchain_store(entry, frame.pc)) + break; + } while (sb1_unwind_user(&frame)); + current->thread.cp0_baduaddr = saved_baduaddr; +} static void save_raw_perf_callchain(struct perf_callchain_entry_ctx *entry, unsigned long reg29) -- 2.53.0