From nobody Fri Nov 29 21:55:48 2024 Received: from mx0a-001b2d01.pphosted.com (mx0a-001b2d01.pphosted.com [148.163.156.1]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id F343E1494C9; Sun, 15 Sep 2024 20:58:12 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=148.163.156.1 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1726433895; cv=none; b=CM0uLmNH5xEq2JDQDaXPsbc1twaJjE29WRu6wP+vsnj0Po1rhBfeR7UirS0zyi7+ijkSSF2aQLEmAj7vB6OEtioDzSzAFBWG5wmgqfcFsLIkqyctEm+uTqVnBusWm++wDWiycWoMG+ISmrXVznPZ+cNI8wpzVYYwnfMcsuEf+Bk= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1726433895; c=relaxed/simple; bh=X3d9sojtvyqYoHi5pmCO8Tm9RSUBDrvSYL+9x3vSvRE=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=Uh1TPL3YHgjg9k32WHeXMyVoOseR+0f4UwcJqAxnnNgFnsb4gNo26KspZsnKL9C3+J82IgcGYW2X19gWngoABdFO59SfDrOjD9lB3EZe1IWqid13d/J+nqiz6/c1u4Mch+WjV/Ol6KgcozW41DepV79EfTM7IRF50YJbLrZsZCc= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.ibm.com; spf=pass smtp.mailfrom=linux.ibm.com; dkim=pass (2048-bit key) header.d=ibm.com header.i=@ibm.com header.b=X88UKZgi; arc=none smtp.client-ip=148.163.156.1 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.ibm.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.ibm.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=ibm.com header.i=@ibm.com header.b="X88UKZgi" Received: from pps.filterd (m0353729.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.18.1.2/8.18.1.2) with ESMTP id 48FFobD8025501; Sun, 15 Sep 2024 20:57:51 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ibm.com; h=from :to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; s=pp1; bh=rqOyCr0uCfPwq ckG0LZsDvWReQOYVR8VwLXpWx2uZUY=; b=X88UKZgiprM27XCphbtfInmUhuoOb wX/x04Iov2jxEP3sxmy6DvvnbRvcsUcG3F5KlblIpBcJX8jfZN10OQA/0zNjHwI9 yrH5QtK+ualdngOVv1Qci+9baL11G1R/GADcZ14dJJvLzQGmipWED9U9zvD97Pmk a1Zc71WSWfX3Usi69x0kLU8AeH/+AvLfK6hCmUWYs5lUpxNravijW5xC4v2BJm/L PcPP6Q7L8vOzJ02A9nadaJT23ElxQ/XgAt72zGS3CyYHqF1C1+rVCwYzToUkcfkJ fgdyqiDDCE8b7ijcnnkOIgLGEew4Zz8qcXZ+bCHmd6qDCcFjTTPhMDOMg== Received: from pps.reinject (localhost [127.0.0.1]) by mx0a-001b2d01.pphosted.com (PPS) with ESMTPS id 41n3uhxdx9-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Sun, 15 Sep 2024 20:57:50 +0000 (GMT) Received: from m0353729.ppops.net (m0353729.ppops.net [127.0.0.1]) by pps.reinject (8.18.0.8/8.18.0.8) with ESMTP id 48FKvnTI028353; Sun, 15 Sep 2024 20:57:50 GMT Received: from ppma23.wdc07v.mail.ibm.com (5d.69.3da9.ip4.static.sl-reverse.com [169.61.105.93]) by mx0a-001b2d01.pphosted.com (PPS) with ESMTPS id 41n3uhxdx6-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Sun, 15 Sep 2024 20:57:49 +0000 (GMT) Received: from pps.filterd (ppma23.wdc07v.mail.ibm.com [127.0.0.1]) by ppma23.wdc07v.mail.ibm.com (8.18.1.2/8.18.1.2) with ESMTP id 48FKBlkB030649; Sun, 15 Sep 2024 20:57:48 GMT Received: from smtprelay05.fra02v.mail.ibm.com ([9.218.2.225]) by ppma23.wdc07v.mail.ibm.com (PPS) with ESMTPS id 41npamumhb-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=NOT); Sun, 15 Sep 2024 20:57:48 +0000 Received: from smtpav06.fra02v.mail.ibm.com (smtpav06.fra02v.mail.ibm.com [10.20.54.105]) by smtprelay05.fra02v.mail.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 48FKviTS54329656 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Sun, 15 Sep 2024 20:57:44 GMT Received: from smtpav06.fra02v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 3096C20049; Sun, 15 Sep 2024 20:57:44 +0000 (GMT) Received: from smtpav06.fra02v.mail.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 750B620040; Sun, 15 Sep 2024 20:57:40 +0000 (GMT) Received: from li-bd3f974c-2712-11b2-a85c-df1cec4d728e.ibm.com.com (unknown [9.43.68.55]) by smtpav06.fra02v.mail.ibm.com (Postfix) with ESMTP; Sun, 15 Sep 2024 20:57:40 +0000 (GMT) From: Hari Bathini To: linuxppc-dev , bpf@vger.kernel.org, linux-trace-kernel@vger.kernel.org, linux-kbuild@vger.kernel.org, linux-kernel@vger.kernel.org Cc: "Naveen N. Rao" , Mark Rutland , Daniel Borkmann , Masahiro Yamada , Nicholas Piggin , Alexei Starovoitov , Steven Rostedt , Andrii Nakryiko , Christophe Leroy , Vishal Chourasia , Mahesh J Salgaonkar , Masami Hiramatsu Subject: [PATCH v5 12/17] powerpc64/ftrace: Move ftrace sequence out of line Date: Mon, 16 Sep 2024 02:26:43 +0530 Message-ID: <20240915205648.830121-13-hbathini@linux.ibm.com> X-Mailer: git-send-email 2.46.0 In-Reply-To: <20240915205648.830121-1-hbathini@linux.ibm.com> References: <20240915205648.830121-1-hbathini@linux.ibm.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-TM-AS-GCONF: 00 X-Proofpoint-ORIG-GUID: axvE66zg7DbLsFFt0pgVHTG0EDIiuow_ X-Proofpoint-GUID: MvEd_wKlqsj979q4D183eI4bFNUL4PgX X-Proofpoint-Virus-Version: vendor=baseguard engine=ICAP:2.0.293,Aquarius:18.0.1039,Hydra:6.0.680,FMLib:17.12.60.29 definitions=2024-09-15_12,2024-09-13_02,2024-09-02_01 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 priorityscore=1501 mlxlogscore=999 adultscore=0 phishscore=0 spamscore=0 lowpriorityscore=0 clxscore=1015 bulkscore=0 suspectscore=0 mlxscore=0 impostorscore=0 malwarescore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.19.0-2408220000 definitions=main-2409150159 Content-Type: text/plain; charset="utf-8" From: Naveen N Rao Function profile sequence on powerpc includes two instructions at the beginning of each function: mflr r0 bl ftrace_caller The call to ftrace_caller() gets nop'ed out during kernel boot and is patched in when ftrace is enabled. Given the sequence, we cannot return from ftrace_caller with 'blr' as we need to keep LR and r0 intact. This results in link stack (return address predictor) imbalance when ftrace is enabled. To address that, we would like to use a three instruction sequence: mflr r0 bl ftrace_caller mtlr r0 Further more, to support DYNAMIC_FTRACE_WITH_CALL_OPS, we need to reserve two instruction slots before the function. This results in a total of five instruction slots to be reserved for ftrace use on each function that is traced. Move the function profile sequence out-of-line to minimize its impact. To do this, we reserve a single nop at function entry using -fpatchable-function-entry=3D1 and add a pass on vmlinux.o to determine the total number of functions that can be traced. This is then used to generate a .S file reserving the appropriate amount of space for use as ftrace stubs, which is built and linked into vmlinux. On bootup, the stub space is split into separate stubs per function and populated with the proper instruction sequence. A pointer to the associated stub is maintained in dyn_arch_ftrace. For modules, space for ftrace stubs is reserved from the generic module stub space. This is restricted to and enabled by default only on 64-bit powerpc, though there are some changes to accommodate 32-bit powerpc. This is done so that 32-bit powerpc could choose to opt into this based on further tests and benchmarks. As an example, after this patch, kernel functions will have a single nop at function entry: : addis r2,r12,467 addi r2,r2,-16028 nop mfocrf r11,8 ... When ftrace is enabled, the nop is converted to an unconditional branch to the stub associated with that function: : addis r2,r12,467 addi r2,r2,-16028 b ftrace_ool_stub_text_end+0x11b28 mfocrf r11,8 ... The associated stub: : mflr r0 bl ftrace_caller mtlr r0 b kernel_clone+0xc ... This change showed an improvement of ~10% in null_syscall benchmark on a Power 10 system with ftrace enabled. Signed-off-by: Naveen N Rao Signed-off-by: Hari Bathini --- Changes in v5: * Fixed ftrace stack tracer failure due to inadvertent use of 'add r7, r3, MCOUNT_INSN_SIZE' instruction instead of 'addi r7, r3, MCOUNT_INSN_SIZE' * Fixed build error for !CONFIG_MODULES case. * .vmlinux.arch.* files compiled under arch/powerpc/tools * Made sure .vmlinux.arch.* files are cleaned with `make clean` arch/powerpc/Kbuild | 2 +- arch/powerpc/Kconfig | 5 + arch/powerpc/Makefile | 4 + arch/powerpc/include/asm/ftrace.h | 11 ++ arch/powerpc/include/asm/module.h | 5 + arch/powerpc/kernel/asm-offsets.c | 4 + arch/powerpc/kernel/module_64.c | 58 +++++++- arch/powerpc/kernel/trace/ftrace.c | 162 +++++++++++++++++++-- arch/powerpc/kernel/trace/ftrace_entry.S | 116 +++++++++++---- arch/powerpc/tools/Makefile | 12 ++ arch/powerpc/tools/ftrace-gen-ool-stubs.sh | 43 ++++++ 11 files changed, 384 insertions(+), 38 deletions(-) create mode 100644 arch/powerpc/tools/Makefile create mode 100755 arch/powerpc/tools/ftrace-gen-ool-stubs.sh diff --git a/arch/powerpc/Kbuild b/arch/powerpc/Kbuild index 571f260b0842..b010ccb071b6 100644 --- a/arch/powerpc/Kbuild +++ b/arch/powerpc/Kbuild @@ -19,4 +19,4 @@ obj-$(CONFIG_KEXEC_CORE) +=3D kexec/ obj-$(CONFIG_KEXEC_FILE) +=3D purgatory/ =20 # for cleaning -subdir- +=3D boot +subdir- +=3D boot tools diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index de18f3baff66..bae96b65f295 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -568,6 +568,11 @@ config ARCH_USING_PATCHABLE_FUNCTION_ENTRY def_bool $(success,$(srctree)/arch/powerpc/tools/gcc-check-fpatchable-fun= ction-entry.sh $(CC) -mlittle-endian) if PPC64 && CPU_LITTLE_ENDIAN def_bool $(success,$(srctree)/arch/powerpc/tools/gcc-check-fpatchable-fun= ction-entry.sh $(CC) -mbig-endian) if PPC64 && CPU_BIG_ENDIAN =20 +config PPC_FTRACE_OUT_OF_LINE + def_bool PPC64 && ARCH_USING_PATCHABLE_FUNCTION_ENTRY + depends on PPC64 + select ARCH_WANTS_PRE_LINK_VMLINUX + config HOTPLUG_CPU bool "Support for enabling/disabling CPUs" depends on SMP && (PPC_PSERIES || \ diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile index bbfe4a1f06ef..c973e6cd1ae8 100644 --- a/arch/powerpc/Makefile +++ b/arch/powerpc/Makefile @@ -155,7 +155,11 @@ CC_FLAGS_NO_FPU :=3D $(call cc-option,-msoft-float) ifdef CONFIG_FUNCTION_TRACER ifdef CONFIG_ARCH_USING_PATCHABLE_FUNCTION_ENTRY KBUILD_CPPFLAGS +=3D -DCC_USING_PATCHABLE_FUNCTION_ENTRY +ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE +CC_FLAGS_FTRACE :=3D -fpatchable-function-entry=3D1 +else CC_FLAGS_FTRACE :=3D -fpatchable-function-entry=3D2 +endif else CC_FLAGS_FTRACE :=3D -pg ifdef CONFIG_MPROFILE_KERNEL diff --git a/arch/powerpc/include/asm/ftrace.h b/arch/powerpc/include/asm/f= trace.h index 278d4548e8f1..bdbafc668b20 100644 --- a/arch/powerpc/include/asm/ftrace.h +++ b/arch/powerpc/include/asm/ftrace.h @@ -24,6 +24,10 @@ unsigned long prepare_ftrace_return(unsigned long parent= , unsigned long ip, struct module; struct dyn_ftrace; struct dyn_arch_ftrace { +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + /* pointer to the associated out-of-line stub */ + unsigned long ool_stub; +#endif }; =20 #ifdef CONFIG_DYNAMIC_FTRACE_WITH_ARGS @@ -130,6 +134,13 @@ static inline u8 this_cpu_get_ftrace_enabled(void) { r= eturn 1; } =20 #ifdef CONFIG_FUNCTION_TRACER extern unsigned int ftrace_tramp_text[], ftrace_tramp_init[]; +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE +struct ftrace_ool_stub { + u32 insn[4]; +}; +extern struct ftrace_ool_stub ftrace_ool_stub_text_end[], ftrace_ool_stub_= inittext[]; +extern unsigned int ftrace_ool_stub_text_end_count, ftrace_ool_stub_initte= xt_count; +#endif void ftrace_free_init_tramp(void); unsigned long ftrace_call_adjust(unsigned long addr); #else diff --git a/arch/powerpc/include/asm/module.h b/arch/powerpc/include/asm/m= odule.h index 300c777cc307..9ee70a4a0fde 100644 --- a/arch/powerpc/include/asm/module.h +++ b/arch/powerpc/include/asm/module.h @@ -47,6 +47,11 @@ struct mod_arch_specific { #ifdef CONFIG_DYNAMIC_FTRACE unsigned long tramp; unsigned long tramp_regs; +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + struct ftrace_ool_stub *ool_stubs; + unsigned int ool_stub_count; + unsigned int ool_stub_index; +#endif #endif }; =20 diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-of= fsets.c index 23733282de4d..6854547d3164 100644 --- a/arch/powerpc/kernel/asm-offsets.c +++ b/arch/powerpc/kernel/asm-offsets.c @@ -674,5 +674,9 @@ int main(void) DEFINE(BPT_SIZE, BPT_SIZE); #endif =20 +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + DEFINE(FTRACE_OOL_STUB_SIZE, sizeof(struct ftrace_ool_stub)); +#endif + return 0; } diff --git a/arch/powerpc/kernel/module_64.c b/arch/powerpc/kernel/module_6= 4.c index 1db88409bd95..6816e9967cab 100644 --- a/arch/powerpc/kernel/module_64.c +++ b/arch/powerpc/kernel/module_64.c @@ -205,7 +205,9 @@ static int relacmp(const void *_x, const void *_y) =20 /* Get size of potential trampolines required. */ static unsigned long get_stubs_size(const Elf64_Ehdr *hdr, - const Elf64_Shdr *sechdrs) + const Elf64_Shdr *sechdrs, + char *secstrings, + struct module *me) { /* One extra reloc so it's always 0-addr terminated */ unsigned long relocs =3D 1; @@ -244,6 +246,24 @@ static unsigned long get_stubs_size(const Elf64_Ehdr *= hdr, /* stubs for ftrace_caller and ftrace_regs_caller */ relocs +=3D IS_ENABLED(CONFIG_DYNAMIC_FTRACE) + IS_ENABLED(CONFIG_DYNAMIC= _FTRACE_WITH_REGS); =20 +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + /* stubs for the function tracer */ + for (i =3D 1; i < hdr->e_shnum; i++) { + if (!strcmp(secstrings + sechdrs[i].sh_name, "__patchable_function_entri= es")) { + me->arch.ool_stub_count =3D sechdrs[i].sh_size / sizeof(unsigned long); + me->arch.ool_stub_index =3D 0; + relocs +=3D roundup(me->arch.ool_stub_count * sizeof(struct ftrace_ool_= stub), + sizeof(struct ppc64_stub_entry)) / + sizeof(struct ppc64_stub_entry); + break; + } + } + if (i =3D=3D hdr->e_shnum) { + pr_err("%s: doesn't contain __patchable_function_entries.\n", me->name); + return -ENOEXEC; + } +#endif + pr_debug("Looks like a total of %lu stubs, max\n", relocs); return relocs * sizeof(struct ppc64_stub_entry); } @@ -454,7 +474,7 @@ int module_frob_arch_sections(Elf64_Ehdr *hdr, #endif =20 /* Override the stubs size */ - sechdrs[me->arch.stubs_section].sh_size =3D get_stubs_size(hdr, sechdrs); + sechdrs[me->arch.stubs_section].sh_size =3D get_stubs_size(hdr, sechdrs, = secstrings, me); =20 return 0; } @@ -1079,6 +1099,37 @@ int module_trampoline_target(struct module *mod, uns= igned long addr, return 0; } =20 +static int setup_ftrace_ool_stubs(const Elf64_Shdr *sechdrs, unsigned long= addr, struct module *me) +{ +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + unsigned int i, total_stubs, num_stubs; + struct ppc64_stub_entry *stub; + + total_stubs =3D sechdrs[me->arch.stubs_section].sh_size / sizeof(*stub); + num_stubs =3D roundup(me->arch.ool_stub_count * sizeof(struct ftrace_ool_= stub), + sizeof(struct ppc64_stub_entry)) / sizeof(struct ppc64_stub_entry); + + /* Find the next available entry */ + stub =3D (void *)sechdrs[me->arch.stubs_section].sh_addr; + for (i =3D 0; stub_func_addr(stub[i].funcdata); i++) + if (WARN_ON(i >=3D total_stubs)) + return -1; + + if (WARN_ON(i + num_stubs > total_stubs)) + return -1; + + stub +=3D i; + me->arch.ool_stubs =3D (struct ftrace_ool_stub *)stub; + + /* reserve stubs */ + for (i =3D 0; i < num_stubs; i++) + if (patch_u32((void *)&stub->funcdata, PPC_RAW_NOP())) + return -1; +#endif + + return 0; +} + int module_finalize_ftrace(struct module *mod, const Elf_Shdr *sechdrs) { mod->arch.tramp =3D stub_for_addr(sechdrs, @@ -1097,6 +1148,9 @@ int module_finalize_ftrace(struct module *mod, const = Elf_Shdr *sechdrs) if (!mod->arch.tramp) return -ENOENT; =20 + if (setup_ftrace_ool_stubs(sechdrs, mod->arch.tramp, mod)) + return -ENOENT; + return 0; } #endif diff --git a/arch/powerpc/kernel/trace/ftrace.c b/arch/powerpc/kernel/trace= /ftrace.c index 719517265d39..1fee074388cc 100644 --- a/arch/powerpc/kernel/trace/ftrace.c +++ b/arch/powerpc/kernel/trace/ftrace.c @@ -37,7 +37,8 @@ unsigned long ftrace_call_adjust(unsigned long addr) if (addr >=3D (unsigned long)__exittext_begin && addr < (unsigned long)__= exittext_end) return 0; =20 - if (IS_ENABLED(CONFIG_ARCH_USING_PATCHABLE_FUNCTION_ENTRY)) + if (IS_ENABLED(CONFIG_ARCH_USING_PATCHABLE_FUNCTION_ENTRY) && + !IS_ENABLED(CONFIG_PPC_FTRACE_OUT_OF_LINE)) addr +=3D MCOUNT_INSN_SIZE; =20 return addr; @@ -127,11 +128,25 @@ static unsigned long ftrace_lookup_module_stub(unsign= ed long ip, unsigned long a } #endif =20 +static unsigned long ftrace_get_ool_stub(struct dyn_ftrace *rec) +{ +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + return rec->arch.ool_stub; +#else + BUILD_BUG(); +#endif +} + static int ftrace_get_call_inst(struct dyn_ftrace *rec, unsigned long addr= , ppc_inst_t *call_inst) { - unsigned long ip =3D rec->ip; + unsigned long ip; unsigned long stub; =20 + if (IS_ENABLED(CONFIG_PPC_FTRACE_OUT_OF_LINE)) + ip =3D ftrace_get_ool_stub(rec) + MCOUNT_INSN_SIZE; /* second instructio= n in stub */ + else + ip =3D rec->ip; + if (is_offset_in_branch_range(addr - ip)) /* Within range */ stub =3D addr; @@ -142,7 +157,7 @@ static int ftrace_get_call_inst(struct dyn_ftrace *rec,= unsigned long addr, ppc_ stub =3D ftrace_lookup_module_stub(ip, addr); =20 if (!stub) { - pr_err("0x%lx: No ftrace stubs reachable\n", ip); + pr_err("0x%lx (0x%lx): No ftrace stubs reachable\n", ip, rec->ip); return -EINVAL; } =20 @@ -150,6 +165,92 @@ static int ftrace_get_call_inst(struct dyn_ftrace *rec= , unsigned long addr, ppc_ return 0; } =20 +static int ftrace_init_ool_stub(struct module *mod, struct dyn_ftrace *rec) +{ +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + static int ool_stub_text_end_index, ool_stub_inittext_index; + int ret =3D 0, ool_stub_count, *ool_stub_index; + ppc_inst_t inst; + /* + * See ftrace_entry.S if changing the below instruction sequence, as we r= ely on + * decoding the last branch instruction here to recover the correct funct= ion ip. + */ + struct ftrace_ool_stub *ool_stub, ool_stub_template =3D { + .insn =3D { + PPC_RAW_MFLR(_R0), + PPC_RAW_NOP(), /* bl ftrace_caller */ + PPC_RAW_MTLR(_R0), + PPC_RAW_NOP() /* b rec->ip + 4 */ + } + }; + + WARN_ON(rec->arch.ool_stub); + + if (is_kernel_inittext(rec->ip)) { + ool_stub =3D ftrace_ool_stub_inittext; + ool_stub_index =3D &ool_stub_inittext_index; + ool_stub_count =3D ftrace_ool_stub_inittext_count; + } else if (is_kernel_text(rec->ip)) { + ool_stub =3D ftrace_ool_stub_text_end; + ool_stub_index =3D &ool_stub_text_end_index; + ool_stub_count =3D ftrace_ool_stub_text_end_count; +#ifdef CONFIG_MODULES + } else if (mod) { + ool_stub =3D mod->arch.ool_stubs; + ool_stub_index =3D &mod->arch.ool_stub_index; + ool_stub_count =3D mod->arch.ool_stub_count; +#endif + } else { + return -EINVAL; + } + + ool_stub +=3D (*ool_stub_index)++; + + if (WARN_ON(*ool_stub_index > ool_stub_count)) + return -EINVAL; + + if (!is_offset_in_branch_range((long)rec->ip - (long)&ool_stub->insn[0]) = || + !is_offset_in_branch_range((long)(rec->ip + MCOUNT_INSN_SIZE) - + (long)&ool_stub->insn[3])) { + pr_err("%s: ftrace ool stub out of range (%p -> %p).\n", + __func__, (void *)rec->ip, (void *)&ool_stub->insn[0]); + return -EINVAL; + } + + rec->arch.ool_stub =3D (unsigned long)&ool_stub->insn[0]; + + /* bl ftrace_caller */ + if (!mod) + ret =3D ftrace_get_call_inst(rec, (unsigned long)ftrace_caller, &inst); +#ifdef CONFIG_MODULES + else + /* + * We can't use ftrace_get_call_inst() since that uses + * __module_text_address(rec->ip) to look up the module. + * But, since the module is not fully formed at this stage, + * the lookup fails. We know the target though, so generate + * the branch inst directly. + */ + inst =3D ftrace_create_branch_inst(ftrace_get_ool_stub(rec) + MCOUNT_INS= N_SIZE, + mod->arch.tramp, 1); +#endif + ool_stub_template.insn[1] =3D ppc_inst_val(inst); + + /* b rec->ip + 4 */ + if (!ret && create_branch(&inst, &ool_stub->insn[3], rec->ip + MCOUNT_INS= N_SIZE, 0)) + return -EINVAL; + ool_stub_template.insn[3] =3D ppc_inst_val(inst); + + if (!ret) + ret =3D patch_instructions((u32 *)ool_stub, (u32 *)&ool_stub_template, + sizeof(ool_stub_template), false); + + return ret; +#else /* !CONFIG_PPC_FTRACE_OUT_OF_LINE */ + BUILD_BUG(); +#endif +} + #ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr, uns= igned long addr) { @@ -162,18 +263,29 @@ int ftrace_modify_call(struct dyn_ftrace *rec, unsign= ed long old_addr, unsigned int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) { ppc_inst_t old, new; - int ret; + unsigned long ip =3D rec->ip; + int ret =3D 0; =20 /* This can only ever be called during module load */ - if (WARN_ON(!IS_ENABLED(CONFIG_MODULES) || core_kernel_text(rec->ip))) + if (WARN_ON(!IS_ENABLED(CONFIG_MODULES) || core_kernel_text(ip))) return -EINVAL; =20 old =3D ppc_inst(PPC_RAW_NOP()); - ret =3D ftrace_get_call_inst(rec, addr, &new); - if (ret) - return ret; + if (IS_ENABLED(CONFIG_PPC_FTRACE_OUT_OF_LINE)) { + ip =3D ftrace_get_ool_stub(rec) + MCOUNT_INSN_SIZE; /* second instructio= n in stub */ + ret =3D ftrace_get_call_inst(rec, (unsigned long)ftrace_caller, &old); + } + + ret |=3D ftrace_get_call_inst(rec, addr, &new); + + if (!ret) + ret =3D ftrace_modify_code(ip, old, new); =20 - return ftrace_modify_code(rec->ip, old, new); + if (!ret && IS_ENABLED(CONFIG_PPC_FTRACE_OUT_OF_LINE)) + ret =3D ftrace_modify_code(rec->ip, ppc_inst(PPC_RAW_NOP()), + ppc_inst(PPC_RAW_BRANCH((long)ftrace_get_ool_stub(rec) - (long)rec->ip= ))); + + return ret; } =20 int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec, unsigned l= ong addr) @@ -206,6 +318,13 @@ void ftrace_replace_code(int enable) new_addr =3D ftrace_get_addr_new(rec); update =3D ftrace_update_record(rec, enable); =20 + if (IS_ENABLED(CONFIG_PPC_FTRACE_OUT_OF_LINE) && update !=3D FTRACE_UPDA= TE_IGNORE) { + ip =3D ftrace_get_ool_stub(rec) + MCOUNT_INSN_SIZE; + ret =3D ftrace_get_call_inst(rec, (unsigned long)ftrace_caller, &nop_in= st); + if (ret) + goto out; + } + switch (update) { case FTRACE_UPDATE_IGNORE: default: @@ -230,6 +349,24 @@ void ftrace_replace_code(int enable) =20 if (!ret) ret =3D ftrace_modify_code(ip, old, new); + + if (!ret && IS_ENABLED(CONFIG_PPC_FTRACE_OUT_OF_LINE) && + (update =3D=3D FTRACE_UPDATE_MAKE_NOP || update =3D=3D FTRACE_UPDATE= _MAKE_CALL)) { + /* Update the actual ftrace location */ + call_inst =3D ppc_inst(PPC_RAW_BRANCH((long)ftrace_get_ool_stub(rec) - + (long)rec->ip)); + nop_inst =3D ppc_inst(PPC_RAW_NOP()); + ip =3D rec->ip; + + if (update =3D=3D FTRACE_UPDATE_MAKE_NOP) + ret =3D ftrace_modify_code(ip, call_inst, nop_inst); + else + ret =3D ftrace_modify_code(ip, nop_inst, call_inst); + + if (ret) + goto out; + } + if (ret) goto out; } @@ -249,7 +386,8 @@ int ftrace_init_nop(struct module *mod, struct dyn_ftra= ce *rec) /* Verify instructions surrounding the ftrace location */ if (IS_ENABLED(CONFIG_ARCH_USING_PATCHABLE_FUNCTION_ENTRY)) { /* Expect nops */ - ret =3D ftrace_validate_inst(ip - 4, ppc_inst(PPC_RAW_NOP())); + if (!IS_ENABLED(CONFIG_PPC_FTRACE_OUT_OF_LINE)) + ret =3D ftrace_validate_inst(ip - 4, ppc_inst(PPC_RAW_NOP())); if (!ret) ret =3D ftrace_validate_inst(ip, ppc_inst(PPC_RAW_NOP())); } else if (IS_ENABLED(CONFIG_PPC32)) { @@ -277,6 +415,10 @@ int ftrace_init_nop(struct module *mod, struct dyn_ftr= ace *rec) if (ret) return ret; =20 + /* Set up out-of-line stub */ + if (IS_ENABLED(CONFIG_PPC_FTRACE_OUT_OF_LINE)) + return ftrace_init_ool_stub(mod, rec); + /* Nop-out the ftrace location */ new =3D ppc_inst(PPC_RAW_NOP()); addr =3D MCOUNT_ADDR; diff --git a/arch/powerpc/kernel/trace/ftrace_entry.S b/arch/powerpc/kernel= /trace/ftrace_entry.S index 244a1c7bb1e8..5b2fc6483dce 100644 --- a/arch/powerpc/kernel/trace/ftrace_entry.S +++ b/arch/powerpc/kernel/trace/ftrace_entry.S @@ -56,7 +56,7 @@ SAVE_GPR(2, r1) SAVE_GPRS(11, 31, r1) .else -#ifdef CONFIG_LIVEPATCH_64 +#if defined(CONFIG_LIVEPATCH_64) || defined(CONFIG_PPC_FTRACE_OUT_OF_LINE) SAVE_GPR(14, r1) #endif .endif @@ -78,10 +78,6 @@ =20 /* Get the _mcount() call site out of LR */ mflr r7 - /* Save it as pt_regs->nip */ - PPC_STL r7, _NIP(r1) - /* Also save it in B's stackframe header for proper unwind */ - PPC_STL r7, LRSAVE+SWITCH_FRAME_SIZE(r1) /* Save the read LR in pt_regs->link */ PPC_STL r0, _LINK(r1) =20 @@ -96,16 +92,6 @@ lwz r5,function_trace_op@l(r3) #endif =20 -#ifdef CONFIG_LIVEPATCH_64 - mr r14, r7 /* remember old NIP */ -#endif - - /* Calculate ip from nip-4 into r3 for call below */ - subi r3, r7, MCOUNT_INSN_SIZE - - /* Put the original return address in r4 as parent_ip */ - mr r4, r0 - /* Save special regs */ PPC_STL r8, _MSR(r1) .if \allregs =3D=3D 1 @@ -114,17 +100,69 @@ PPC_STL r11, _CCR(r1) .endif =20 +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + /* Save our real return address in nvr for return */ + .if \allregs =3D=3D 0 + SAVE_GPR(15, r1) + .endif + mr r15, r7 + /* + * We want the ftrace location in the function, but our lr (in r7) + * points at the 'mtlr r0' instruction in the out of line stub. To + * recover the ftrace location, we read the branch instruction in the + * stub, and adjust our lr by the branch offset. + * + * See ftrace_init_ool_stub() for the profile sequence. + */ + lwz r8, MCOUNT_INSN_SIZE(r7) + slwi r8, r8, 6 + srawi r8, r8, 6 + add r3, r7, r8 + /* + * Override our nip to point past the branch in the original function. + * This allows reliable stack trace and the ftrace stack tracer to work a= s-is. + */ + addi r7, r3, MCOUNT_INSN_SIZE +#else + /* Calculate ip from nip-4 into r3 for call below */ + subi r3, r7, MCOUNT_INSN_SIZE +#endif + + /* Save NIP as pt_regs->nip */ + PPC_STL r7, _NIP(r1) + /* Also save it in B's stackframe header for proper unwind */ + PPC_STL r7, LRSAVE+SWITCH_FRAME_SIZE(r1) +#if defined(CONFIG_LIVEPATCH_64) || defined(CONFIG_PPC_FTRACE_OUT_OF_LINE) + mr r14, r7 /* remember old NIP */ +#endif + + /* Put the original return address in r4 as parent_ip */ + mr r4, r0 + /* Load &pt_regs in r6 for call below */ addi r6, r1, STACK_INT_FRAME_REGS .endm =20 .macro ftrace_regs_exit allregs +#ifndef CONFIG_PPC_FTRACE_OUT_OF_LINE /* Load ctr with the possibly modified NIP */ PPC_LL r3, _NIP(r1) mtctr r3 =20 #ifdef CONFIG_LIVEPATCH_64 cmpd r14, r3 /* has NIP been altered? */ +#endif +#else /* !CONFIG_PPC_FTRACE_OUT_OF_LINE */ + /* Load LR with the possibly modified NIP */ + PPC_LL r3, _NIP(r1) + cmpd r14, r3 /* has NIP been altered? */ + bne- 1f + + mr r3, r15 + .if \allregs =3D=3D 0 + REST_GPR(15, r1) + .endif +1: mtlr r3 #endif =20 /* Restore gprs */ @@ -132,14 +170,16 @@ REST_GPRS(2, 31, r1) .else REST_GPRS(3, 10, r1) -#ifdef CONFIG_LIVEPATCH_64 +#if defined(CONFIG_LIVEPATCH_64) || defined(CONFIG_PPC_FTRACE_OUT_OF_LINE) REST_GPR(14, r1) #endif .endif =20 /* Restore possibly modified LR */ PPC_LL r0, _LINK(r1) +#ifndef CONFIG_PPC_FTRACE_OUT_OF_LINE mtlr r0 +#endif =20 #ifdef CONFIG_PPC64 /* Restore callee's TOC */ @@ -153,7 +193,16 @@ /* Based on the cmpd above, if the NIP was altered handle livepatc= h */ bne- livepatch_handler #endif - bctr /* jump after _mcount site */ + /* jump after _mcount site */ +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + /* + * Return with blr to keep the link stack balanced. The function profilin= g sequence + * uses 'mtlr r0' to restore LR. + */ + blr +#else + bctr +#endif .endm =20 _GLOBAL(ftrace_regs_caller) @@ -177,6 +226,11 @@ _GLOBAL(ftrace_stub) =20 #ifdef CONFIG_PPC64 ftrace_no_trace: +#ifdef CONFIG_PPC_FTRACE_OUT_OF_LINE + REST_GPR(3, r1) + addi r1, r1, SWITCH_FRAME_SIZE+STACK_FRAME_MIN_SIZE + blr +#else mflr r3 mtctr r3 REST_GPR(3, r1) @@ -184,6 +238,7 @@ ftrace_no_trace: mtlr r0 bctr #endif +#endif =20 #ifdef CONFIG_LIVEPATCH_64 /* @@ -194,11 +249,17 @@ ftrace_no_trace: * We get here when a function A, calls another function B, but B has * been live patched with a new function C. * - * On entry: - * - we have no stack frame and can not allocate one + * On entry, we have no stack frame and can not allocate one. + * + * With PPC_FTRACE_OUT_OF_LINE=3Dn, on entry: * - LR points back to the original caller (in A) * - CTR holds the new NIP in C * - r0, r11 & r12 are free + * + * With PPC_FTRACE_OUT_OF_LINE=3Dy, on entry: + * - r0 points back to the original caller (in A) + * - LR holds the new NIP in C + * - r11 & r12 are free */ livepatch_handler: ld r12, PACA_THREAD_INFO(r13) @@ -208,18 +269,23 @@ livepatch_handler: addi r11, r11, 24 std r11, TI_livepatch_sp(r12) =20 - /* Save toc & real LR on livepatch stack */ - std r2, -24(r11) - mflr r12 - std r12, -16(r11) - /* Store stack end marker */ lis r12, STACK_END_MAGIC@h ori r12, r12, STACK_END_MAGIC@l std r12, -8(r11) =20 - /* Put ctr in r12 for global entry and branch there */ + /* Save toc & real LR on livepatch stack */ + std r2, -24(r11) +#ifndef CONFIG_PPC_FTRACE_OUT_OF_LINE + mflr r12 + std r12, -16(r11) mfctr r12 +#else + std r0, -16(r11) + mflr r12 + /* Put ctr in r12 for global entry and branch there */ + mtctr r12 +#endif bctrl =20 /* diff --git a/arch/powerpc/tools/Makefile b/arch/powerpc/tools/Makefile new file mode 100644 index 000000000000..3a389526498e --- /dev/null +++ b/arch/powerpc/tools/Makefile @@ -0,0 +1,12 @@ +# SPDX-License-Identifier: GPL-2.0-or-later + +quiet_cmd_gen_ftrace_ool_stubs =3D GEN $@ + cmd_gen_ftrace_ool_stubs =3D $< vmlinux.o $@ + +$(obj)/.vmlinux.arch.S: $(src)/ftrace-gen-ool-stubs.sh vmlinux.o FORCE + $(call if_changed,gen_ftrace_ool_stubs) + +$(obj)/.vmlinux.arch.o: $(obj)/.vmlinux.arch.S FORCE + $(call if_changed_rule,as_o_S) + +clean-files +=3D .vmlinux.arch.S .vmlinux.arch.o diff --git a/arch/powerpc/tools/ftrace-gen-ool-stubs.sh b/arch/powerpc/tool= s/ftrace-gen-ool-stubs.sh new file mode 100755 index 000000000000..8e0a6d4ea202 --- /dev/null +++ b/arch/powerpc/tools/ftrace-gen-ool-stubs.sh @@ -0,0 +1,43 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0-or-later + +# Error out on error +set -e + +is_enabled() { + grep -q "^$1=3Dy" include/config/auto.conf +} + +vmlinux_o=3D${1} +arch_vmlinux_S=3D${2} + +RELOCATION=3DR_PPC64_ADDR64 +if is_enabled CONFIG_PPC32; then + RELOCATION=3DR_PPC_ADDR32 +fi + +num_ool_stubs_text=3D$(${CROSS_COMPILE}objdump -r -j __patchable_function_= entries ${vmlinux_o} | + grep -v ".init.text" | grep "${RELOCATION}" | wc -l) +num_ool_stubs_inittext=3D$(${CROSS_COMPILE}objdump -r -j __patchable_funct= ion_entries ${vmlinux_o} | + grep ".init.text" | grep "${RELOCATION}" | wc -l) + +cat > ${arch_vmlinux_S} < +#include + +.pushsection .tramp.ftrace.text,"aw" +SYM_DATA(ftrace_ool_stub_text_end_count, .long ${num_ool_stubs_text}) + +SYM_CODE_START(ftrace_ool_stub_text_end) + .space ${num_ool_stubs_text} * FTRACE_OOL_STUB_SIZE +SYM_CODE_END(ftrace_ool_stub_text_end) +.popsection + +.pushsection .tramp.ftrace.init,"aw" +SYM_DATA(ftrace_ool_stub_inittext_count, .long ${num_ool_stubs_inittext}) + +SYM_CODE_START(ftrace_ool_stub_inittext) + .space ${num_ool_stubs_inittext} * FTRACE_OOL_STUB_SIZE +SYM_CODE_END(ftrace_ool_stub_inittext) +.popsection +EOF --=20 2.46.0