From 6a6fac11b11299aa5bd8532ea863fc2f652af2b6 Mon Sep 17 00:00:00 2001 From: Thomas Richter Date: Tue, 8 Oct 2019 11:38:41 +0200 Subject: [PATCH 01/13] perf jvmti: Link against tools/lib/ctype.h to have weak strlcpy() The build of file libperf-jvmti.so succeeds but the resulting object fails to load: # ~/linux/tools/perf/perf record -k mono -- java \ -XX:+PreserveFramePointer \ -agentpath:/root/linux/tools/perf/libperf-jvmti.so \ hog 100000 123450 Error occurred during initialization of VM Could not find agent library /root/linux/tools/perf/libperf-jvmti.so in absolute path, with error: /root/linux/tools/perf/libperf-jvmti.so: undefined symbol: _ctype Add the missing _ctype symbol into the build script. Fixes: 79743bc927f6 ("perf jvmti: Link against tools/lib/string.o to have weak strlcpy()") Signed-off-by: Thomas Richter Cc: Heiko Carstens Cc: Vasily Gorbik Link: http://lore.kernel.org/lkml/20191008093841.59387-1-tmricht@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/jvmti/Build | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/tools/perf/jvmti/Build b/tools/perf/jvmti/Build index 1e148bbdf820..202cadaaf097 100644 --- a/tools/perf/jvmti/Build +++ b/tools/perf/jvmti/Build @@ -2,7 +2,7 @@ jvmti-y += libjvmti.o jvmti-y += jvmti_agent.o # For strlcpy -jvmti-y += libstring.o +jvmti-y += libstring.o libctype.o CFLAGS_jvmti = -fPIC -DPIC -I$(JDIR)/include -I$(JDIR)/include/linux CFLAGS_REMOVE_jvmti = -Wmissing-declarations @@ -15,3 +15,7 @@ CFLAGS_libstring.o += -Wno-unused-parameter -DETC_PERFCONFIG="BUILD_STR($(ETC_PE $(OUTPUT)jvmti/libstring.o: ../lib/string.c FORCE $(call rule_mkdir) $(call if_changed_dep,cc_o_c) + +$(OUTPUT)jvmti/libctype.o: ../lib/ctype.c FORCE + $(call rule_mkdir) + $(call if_changed_dep,cc_o_c) From 98a8b2e60c69927b1f405c3b001a1de3f4e53901 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Fri, 11 Oct 2019 11:21:40 -0700 Subject: [PATCH 02/13] perf evlist: Fix fix for freed id arrays In the earlier fix for the memory overrun of id arrays I managed to typo the wrong event in the fix. Of course we need to close the current event in the loop, not the original failing event. The same test case as in the original patch still passes. Fixes: 7834fa948beb ("perf evlist: Fix access of freed id arrays") Signed-off-by: Andi Kleen Cc: Jiri Olsa Link: http://lore.kernel.org/lkml/20191011182140.8353-2-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/evlist.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c index d277a98e62df..de79c735e441 100644 --- a/tools/perf/util/evlist.c +++ b/tools/perf/util/evlist.c @@ -1659,7 +1659,7 @@ struct evsel *perf_evlist__reset_weak_group(struct evlist *evsel_list, is_open = false; if (c2->leader == leader) { if (is_open) - perf_evsel__close(&evsel->core); + perf_evsel__close(&c2->core); c2->leader = c2; c2->core.nr_members = 0; } From 6080728ff8e9c9116e52e6f840152356ac2fea56 Mon Sep 17 00:00:00 2001 From: Yunfeng Ye Date: Tue, 15 Oct 2019 16:30:08 +0800 Subject: [PATCH 03/13] perf tools: Fix resource leak of closedir() on the error paths Both build_mem_topology() and rm_rf_depth_pat() have resource leaks of closedir() on the error paths. Fix this by calling closedir() before function returns. Fixes: e2091cedd51b ("perf tools: Add MEM_TOPOLOGY feature to perf data file") Fixes: cdb6b0235f17 ("perf tools: Add pattern name checking to rm_rf") Signed-off-by: Yunfeng Ye Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Alexei Starovoitov Cc: Alexey Budankov Cc: Andi Kleen Cc: Daniel Borkmann Cc: Feilong Lin Cc: Hu Shiyuan Cc: Igor Lubashev Cc: Kan Liang Cc: Mark Rutland Cc: Martin KaFai Lau Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Song Liu Cc: Yonghong Song Link: http://lore.kernel.org/lkml/cd5f7cd2-b80d-6add-20a1-32f4f43e0744@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/header.c | 4 +++- tools/perf/util/util.c | 6 ++++-- 2 files changed, 7 insertions(+), 3 deletions(-) diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 86d9396cb131..becc2d109423 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -1296,8 +1296,10 @@ static int build_mem_topology(struct memory_node *nodes, u64 size, u64 *cntp) continue; if (WARN_ONCE(cnt >= size, - "failed to write MEM_TOPOLOGY, way too many nodes\n")) + "failed to write MEM_TOPOLOGY, way too many nodes\n")) { + closedir(dir); return -1; + } ret = memory_node__read(&nodes[cnt++], idx); } diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c index 5eda6e19c947..ae56c766eda1 100644 --- a/tools/perf/util/util.c +++ b/tools/perf/util/util.c @@ -154,8 +154,10 @@ static int rm_rf_depth_pat(const char *path, int depth, const char **pat) if (!strcmp(d->d_name, ".") || !strcmp(d->d_name, "..")) continue; - if (!match_pat(d->d_name, pat)) - return -2; + if (!match_pat(d->d_name, pat)) { + ret = -2; + break; + } scnprintf(namebuf, sizeof(namebuf), "%s/%s", path, d->d_name); From f948eb45e3af9fb18a0487d0797a773897ef6929 Mon Sep 17 00:00:00 2001 From: "Gustavo A. R. Silva" Date: Mon, 14 Oct 2019 12:10:47 -0500 Subject: [PATCH 04/13] perf annotate: Fix multiple memory and file descriptor leaks Store SYMBOL_ANNOTATE_ERRNO__BPF_MISSING_BTF in variable *ret*, instead of returning in the middle of the function and leaking multiple resources: prog_linfo, btf, s and bfdf. Addresses-Coverity-ID: 1454832 ("Structurally dead code") Fixes: 11aad897f6d1 ("perf annotate: Don't return -1 for error when doing BPF disassembly") Signed-off-by: Gustavo A. R. Silva Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20191014171047.GA30850@embeddedor Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/annotate.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index 4036c7f7b0fb..e42bf572358c 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -1758,7 +1758,7 @@ static int symbol__disassemble_bpf(struct symbol *sym, info_node = perf_env__find_bpf_prog_info(dso->bpf_prog.env, dso->bpf_prog.id); if (!info_node) { - return SYMBOL_ANNOTATE_ERRNO__BPF_MISSING_BTF; + ret = SYMBOL_ANNOTATE_ERRNO__BPF_MISSING_BTF; goto out; } info_linear = info_node->info_linear; From 5a0baf5123236362fda4e9772cc63b7faa16a0df Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Mon, 7 Oct 2019 10:02:21 +0300 Subject: [PATCH 05/13] perf tools: Fix mode setting in copyfile_mode_ns() slow_copyfile() opens the file by name, so "write" permissions must not be removed in copyfile_mode_ns() before calling slow_copyfile(). Example: Before: $ sudo chmod +r /proc/kcore $ sudo setcap "cap_sys_admin,cap_sys_ptrace,cap_syslog,cap_sys_rawio=ep" tools/perf/perf $ tools/perf/perf buildid-cache -k /proc/kcore Couldn't add /proc/kcore After: $ sudo chmod +r /proc/kcore $ sudo setcap "cap_sys_admin,cap_sys_ptrace,cap_syslog,cap_sys_rawio=ep" tools/perf/perf $ tools/perf/perf buildid-cache -v -k /proc/kcore kcore added to build-id cache directory /home/ahunter/.debug/[kernel.kcore]/37e340b1b5a7cf4f57ba8de2bc777359588a957f/2019100709562289 Signed-off-by: Adrian Hunter Acked-by: Jiri Olsa Tested-by: Arnaldo Carvalho de Melo Link: http://lore.kernel.org/lkml/20191007070221.11158-1-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/copyfile.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/tools/perf/util/copyfile.c b/tools/perf/util/copyfile.c index 3fa0db136667..47e03de7c235 100644 --- a/tools/perf/util/copyfile.c +++ b/tools/perf/util/copyfile.c @@ -101,14 +101,16 @@ static int copyfile_mode_ns(const char *from, const char *to, mode_t mode, if (tofd < 0) goto out; - if (fchmod(tofd, mode)) - goto out_close_to; - if (st.st_size == 0) { /* /proc? do it slowly... */ err = slow_copyfile(from, tmp, nsi); + if (!err && fchmod(tofd, mode)) + err = -1; goto out_close_to; } + if (fchmod(tofd, mode)) + goto out_close_to; + nsinfo__mountns_enter(nsi, &nsc); fromfd = open(from, O_RDONLY); nsinfo__mountns_exit(&nsc); From ae199c580da1754a2b051321eeb76d6dacd8707b Mon Sep 17 00:00:00 2001 From: Yunfeng Ye Date: Tue, 15 Oct 2019 10:54:14 +0800 Subject: [PATCH 06/13] perf c2c: Fix memory leak in build_cl_output() There is a memory leak problem in the failure paths of build_cl_output(), so fix it. Signed-off-by: Yunfeng Ye Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Feilong Lin Cc: Hu Shiyuan Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/4d3c0178-5482-c313-98e1-f82090d2d456@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-c2c.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/tools/perf/builtin-c2c.c b/tools/perf/builtin-c2c.c index 3542b6ab9813..e69f44941aad 100644 --- a/tools/perf/builtin-c2c.c +++ b/tools/perf/builtin-c2c.c @@ -2635,6 +2635,7 @@ static int build_cl_output(char *cl_sort, bool no_source) bool add_sym = false; bool add_dso = false; bool add_src = false; + int ret = 0; if (!buf) return -ENOMEM; @@ -2653,7 +2654,8 @@ static int build_cl_output(char *cl_sort, bool no_source) add_dso = true; } else if (strcmp(tok, "offset")) { pr_err("unrecognized sort token: %s\n", tok); - return -EINVAL; + ret = -EINVAL; + goto err; } } @@ -2676,13 +2678,15 @@ static int build_cl_output(char *cl_sort, bool no_source) add_sym ? "symbol," : "", add_dso ? "dso," : "", add_src ? "cl_srcline," : "", - "node") < 0) - return -ENOMEM; + "node") < 0) { + ret = -ENOMEM; + goto err; + } c2c.show_src = add_src; - +err: free(buf); - return 0; + return ret; } static int setup_coalesce(const char *coalesce, bool no_source) From 7a12f514c408cc499e61211f234858d3d4b7f4ea Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 15 Oct 2019 12:22:37 -0300 Subject: [PATCH 07/13] tools headers kvm: Sync kvm headers with the kernel sources To pick the changes in: bf653b78f960 ("KVM: vmx: Introduce handle_unexpected_vmexit and handle WAITPKG vmexit") That trigger these changes in tooling: CC /tmp/build/perf/arch/x86/util/kvm-stat.o INSTALL GTK UI DESCEND plugins make[3]: Nothing to be done for '/tmp/build/perf/plugins/libtraceevent-dynamic-list'. INSTALL trace_plugins LD /tmp/build/perf/arch/x86/util/perf-in.o LD /tmp/build/perf/arch/x86/perf-in.o LD /tmp/build/perf/arch/perf-in.o LD /tmp/build/perf/perf-in.o LINK /tmp/build/perf/perf And this is not just because that header is included, kvm-stat.c uses the VMX_EXIT_REASONS define and it got changed by the above cset. And addresses this perf build warnings: Warning: Kernel ABI header at 'tools/arch/x86/include/uapi/asm/vmx.h' differs from latest version at 'arch/x86/include/uapi/asm/vmx.h' diff -u tools/arch/x86/include/uapi/asm/vmx.h arch/x86/include/uapi/asm/vmx.h Cc: Adrian Hunter Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Paolo Bonzini Cc: Tao Xu Cc: Wang Nan Link: https://lkml.kernel.org/n/tip-gr1eel0hckmi5l3p2ewdpfxh@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/arch/x86/include/uapi/asm/vmx.h | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/tools/arch/x86/include/uapi/asm/vmx.h b/tools/arch/x86/include/uapi/asm/vmx.h index f01950aa7fae..3eb8411ab60e 100644 --- a/tools/arch/x86/include/uapi/asm/vmx.h +++ b/tools/arch/x86/include/uapi/asm/vmx.h @@ -86,6 +86,8 @@ #define EXIT_REASON_PML_FULL 62 #define EXIT_REASON_XSAVES 63 #define EXIT_REASON_XRSTORS 64 +#define EXIT_REASON_UMWAIT 67 +#define EXIT_REASON_TPAUSE 68 #define VMX_EXIT_REASONS \ { EXIT_REASON_EXCEPTION_NMI, "EXCEPTION_NMI" }, \ @@ -144,7 +146,9 @@ { EXIT_REASON_RDSEED, "RDSEED" }, \ { EXIT_REASON_PML_FULL, "PML_FULL" }, \ { EXIT_REASON_XSAVES, "XSAVES" }, \ - { EXIT_REASON_XRSTORS, "XRSTORS" } + { EXIT_REASON_XRSTORS, "XRSTORS" }, \ + { EXIT_REASON_UMWAIT, "UMWAIT" }, \ + { EXIT_REASON_TPAUSE, "TPAUSE" } #define VMX_ABORT_SAVE_GUEST_MSR_FAIL 1 #define VMX_ABORT_LOAD_HOST_PDPTE_FAIL 2 From 7cb3a2445705e45af9b58dd241d26598a35b1fdd Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 15 Oct 2019 12:30:08 -0300 Subject: [PATCH 08/13] tools headers kvm: Sync kvm headers with the kernel sources To pick the changes in: 0cb8410b90e7 ("kvm: svm: Intercept RDPRU") That trigger a rebuild in too in tooling: CC /tmp/build/perf/arch/x86/util/kvm-stat.o But this time around no changes in tooling results, as SVM_EXIT_RDPRU wasn't added to SVM_EXIT_REASONS, that is used in kvm-stat.c. And addresses this perf build warnings: Warning: Kernel ABI header at 'tools/arch/x86/include/uapi/asm/svm.h' differs from latest version at 'arch/x86/include/uapi/asm/svm.h' diff -u tools/arch/x86/include/uapi/asm/svm.h arch/x86/include/uapi/asm/svm.h Cc: Adrian Hunter Cc: Jim Mattson Cc: Jiri Olsa Cc: Namhyung Kim Cc: Paolo Bonzini Link: https://lkml.kernel.org/n/tip-pqzkt1hmfpqph3ts8i6zzmim@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/arch/x86/include/uapi/asm/svm.h | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/arch/x86/include/uapi/asm/svm.h b/tools/arch/x86/include/uapi/asm/svm.h index a9731f8a480f..2e8a30f06c74 100644 --- a/tools/arch/x86/include/uapi/asm/svm.h +++ b/tools/arch/x86/include/uapi/asm/svm.h @@ -75,6 +75,7 @@ #define SVM_EXIT_MWAIT 0x08b #define SVM_EXIT_MWAIT_COND 0x08c #define SVM_EXIT_XSETBV 0x08d +#define SVM_EXIT_RDPRU 0x08e #define SVM_EXIT_NPF 0x400 #define SVM_EXIT_AVIC_INCOMPLETE_IPI 0x401 #define SVM_EXIT_AVIC_UNACCELERATED_ACCESS 0x402 From 8daf1fb73295b43fb2f624f709449b484532ba64 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 15 Oct 2019 12:35:02 -0300 Subject: [PATCH 09/13] tools headers kvm: Sync kvm.h headers with the kernel sources To pick the changes in: 344c6c804703 ("KVM/Hyper-V: Add new KVM capability KVM_CAP_HYPERV_DIRECT_TLBFLUSH") dee04eee9182 ("KVM: RISC-V: Add KVM_REG_RISCV for ONE_REG interface") These trigger the rebuild of this object: CC /tmp/build/perf/trace/beauty/ioctl.o But do not result in any change in tooling, as the additions are not being used in any table generatator. This silences this perf build warning: Warning: Kernel ABI header at 'tools/include/uapi/linux/kvm.h' differs from latest version at 'include/uapi/linux/kvm.h' diff -u tools/include/uapi/linux/kvm.h include/uapi/linux/kvm.h Cc: Adrian Hunter Cc: Anup Patel Cc: Jiri Olsa Cc: Namhyung Kim Cc: Paolo Bonzini Cc: Paul Walmsley Cc: Tianyu Lan Link: https://lkml.kernel.org/n/tip-d1v48a0qfoe98u5v9tn3mu5u@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/linux/kvm.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/include/uapi/linux/kvm.h b/tools/include/uapi/linux/kvm.h index 233efbb1c81c..52641d8ca9e8 100644 --- a/tools/include/uapi/linux/kvm.h +++ b/tools/include/uapi/linux/kvm.h @@ -999,6 +999,7 @@ struct kvm_ppc_resize_hpt { #define KVM_CAP_ARM_PTRAUTH_GENERIC 172 #define KVM_CAP_PMU_EVENT_FILTER 173 #define KVM_CAP_ARM_IRQ_LINE_LAYOUT_2 174 +#define KVM_CAP_HYPERV_DIRECT_TLBFLUSH 175 #ifdef KVM_CAP_IRQ_ROUTING @@ -1145,6 +1146,7 @@ struct kvm_dirty_tlb { #define KVM_REG_S390 0x5000000000000000ULL #define KVM_REG_ARM64 0x6000000000000000ULL #define KVM_REG_MIPS 0x7000000000000000ULL +#define KVM_REG_RISCV 0x8000000000000000ULL #define KVM_REG_SIZE_SHIFT 52 #define KVM_REG_SIZE_MASK 0x00f0000000000000ULL From 5eca1379c0eb06e3908179665230a86d19bd2df7 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 15 Oct 2019 12:44:00 -0300 Subject: [PATCH 10/13] tools headers UAPI: Sync sched.h with the kernel To get the changes in: 78f6face5af3 ("sched: add kernel-doc for struct clone_args") f14c234b4bc5 ("clone3: switch to copy_struct_from_user()") This file gets rebuilt, but no changes ensues: CC /tmp/build/perf/trace/beauty/clone.o This addresses this perf build warning: Warning: Kernel ABI header at 'tools/include/uapi/linux/sched.h' differs from latest version at 'include/uapi/linux/sched.h' diff -u tools/include/uapi/linux/sched.h include/uapi/linux/sched.h Cc: Adrian Hunter Cc: Aleksa Sarai Cc: Christian Brauner Cc: Jiri Olsa Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-xqruu8wohwlbc57udg1g0xzx@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/linux/sched.h | 30 ++++++++++++++++++++++++++++-- 1 file changed, 28 insertions(+), 2 deletions(-) diff --git a/tools/include/uapi/linux/sched.h b/tools/include/uapi/linux/sched.h index b3105ac1381a..99335e1f4a27 100644 --- a/tools/include/uapi/linux/sched.h +++ b/tools/include/uapi/linux/sched.h @@ -33,8 +33,31 @@ #define CLONE_NEWNET 0x40000000 /* New network namespace */ #define CLONE_IO 0x80000000 /* Clone io context */ -/* - * Arguments for the clone3 syscall +#ifndef __ASSEMBLY__ +/** + * struct clone_args - arguments for the clone3 syscall + * @flags: Flags for the new process as listed above. + * All flags are valid except for CSIGNAL and + * CLONE_DETACHED. + * @pidfd: If CLONE_PIDFD is set, a pidfd will be + * returned in this argument. + * @child_tid: If CLONE_CHILD_SETTID is set, the TID of the + * child process will be returned in the child's + * memory. + * @parent_tid: If CLONE_PARENT_SETTID is set, the TID of + * the child process will be returned in the + * parent's memory. + * @exit_signal: The exit_signal the parent process will be + * sent when the child exits. + * @stack: Specify the location of the stack for the + * child process. + * @stack_size: The size of the stack for the child process. + * @tls: If CLONE_SETTLS is set, the tls descriptor + * is set to tls. + * + * The structure is versioned by size and thus extensible. + * New struct members must go at the end of the struct and + * must be properly 64bit aligned. */ struct clone_args { __aligned_u64 flags; @@ -46,6 +69,9 @@ struct clone_args { __aligned_u64 stack_size; __aligned_u64 tls; }; +#endif + +#define CLONE_ARGS_SIZE_VER0 64 /* sizeof first published struct */ /* * Scheduling policies From 1abecfcaa7bba21c9985e0136fa49836164dd8fd Mon Sep 17 00:00:00 2001 From: Yunfeng Ye Date: Wed, 16 Oct 2019 16:38:45 +0800 Subject: [PATCH 11/13] perf kmem: Fix memory leak in compact_gfp_flags() The memory @orig_flags is allocated by strdup(), it is freed on the normal path, but leak to free on the error path. Fix this by adding free(orig_flags) on the error path. Fixes: 0e11115644b3 ("perf kmem: Print gfp flags in human readable string") Signed-off-by: Yunfeng Ye Cc: Alexander Shishkin Cc: Feilong Lin Cc: Hu Shiyuan Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/f9e9f458-96f3-4a97-a1d5-9feec2420e07@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-kmem.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c index 1e61e353f579..9661671cc26e 100644 --- a/tools/perf/builtin-kmem.c +++ b/tools/perf/builtin-kmem.c @@ -691,6 +691,7 @@ static char *compact_gfp_flags(char *gfp_flags) new = realloc(new_flags, len + strlen(cpt) + 2); if (new == NULL) { free(new_flags); + free(orig_flags); return NULL; } From 13301c6b16a6d809b331bb88e40ab9ce38238b8b Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Sun, 20 Oct 2019 00:07:26 +0200 Subject: [PATCH 12/13] perf/x86/intel/pt: Fix base for single entry topa Jan reported failing ltp test for PT: https://github.com/linux-test-project/ltp/blob/master/testcases/kernel/tracing/pt_test/pt_test.c It looks like the reason is this new commit added in this v5.4 merge window: 38bb8d77d0b9 ("perf/x86/intel/pt: Split ToPA metadata and page layout") which did not keep the TOPA_SHIFT for entry base. Add it back. Reported-by: Jan Stancek Signed-off-by: Jiri Olsa Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Linus Torvalds Cc: Mark Rutland Cc: Michael Petlan Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Fixes: 38bb8d77d0b9 ("perf/x86/intel/pt: Split ToPA metadata and page layout") Link: https://lkml.kernel.org/r/20191019220726.12213-1-jolsa@kernel.org [ Minor changelog edits. ] Signed-off-by: Ingo Molnar --- arch/x86/events/intel/pt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/events/intel/pt.c b/arch/x86/events/intel/pt.c index 74e80ed9c6c4..05e43d0f430b 100644 --- a/arch/x86/events/intel/pt.c +++ b/arch/x86/events/intel/pt.c @@ -627,7 +627,7 @@ static struct topa *topa_alloc(int cpu, gfp_t gfp) * link as the 2nd entry in the table */ if (!intel_pt_validate_hw_cap(PT_CAP_topa_multiple_entries)) { - TOPA_ENTRY(&tp->topa, 1)->base = page_to_phys(p); + TOPA_ENTRY(&tp->topa, 1)->base = page_to_phys(p) >> TOPA_SHIFT; TOPA_ENTRY(&tp->topa, 1)->end = 1; } From 5e6c3c7b1ec217c1c4c95d9148182302b9969b97 Mon Sep 17 00:00:00 2001 From: Thomas Richter Date: Mon, 21 Oct 2019 10:33:54 +0200 Subject: [PATCH 13/13] perf/aux: Fix tracking of auxiliary trace buffer allocation The following commit from the v5.4 merge window: d44248a41337 ("perf/core: Rework memory accounting in perf_mmap()") ... breaks auxiliary trace buffer tracking. If I run command 'perf record -e rbd000' to record samples and saving them in the **auxiliary** trace buffer then the value of 'locked_vm' becomes negative after all trace buffers have been allocated and released: During allocation the values increase: [52.250027] perf_mmap user->locked_vm:0x87 pinned_vm:0x0 ret:0 [52.250115] perf_mmap user->locked_vm:0x107 pinned_vm:0x0 ret:0 [52.250251] perf_mmap user->locked_vm:0x188 pinned_vm:0x0 ret:0 [52.250326] perf_mmap user->locked_vm:0x208 pinned_vm:0x0 ret:0 [52.250441] perf_mmap user->locked_vm:0x289 pinned_vm:0x0 ret:0 [52.250498] perf_mmap user->locked_vm:0x309 pinned_vm:0x0 ret:0 [52.250613] perf_mmap user->locked_vm:0x38a pinned_vm:0x0 ret:0 [52.250715] perf_mmap user->locked_vm:0x408 pinned_vm:0x2 ret:0 [52.250834] perf_mmap user->locked_vm:0x408 pinned_vm:0x83 ret:0 [52.250915] perf_mmap user->locked_vm:0x408 pinned_vm:0x103 ret:0 [52.251061] perf_mmap user->locked_vm:0x408 pinned_vm:0x184 ret:0 [52.251146] perf_mmap user->locked_vm:0x408 pinned_vm:0x204 ret:0 [52.251299] perf_mmap user->locked_vm:0x408 pinned_vm:0x285 ret:0 [52.251383] perf_mmap user->locked_vm:0x408 pinned_vm:0x305 ret:0 [52.251544] perf_mmap user->locked_vm:0x408 pinned_vm:0x386 ret:0 [52.251634] perf_mmap user->locked_vm:0x408 pinned_vm:0x406 ret:0 [52.253018] perf_mmap user->locked_vm:0x408 pinned_vm:0x487 ret:0 [52.253197] perf_mmap user->locked_vm:0x408 pinned_vm:0x508 ret:0 [52.253374] perf_mmap user->locked_vm:0x408 pinned_vm:0x589 ret:0 [52.253550] perf_mmap user->locked_vm:0x408 pinned_vm:0x60a ret:0 [52.253726] perf_mmap user->locked_vm:0x408 pinned_vm:0x68b ret:0 [52.253903] perf_mmap user->locked_vm:0x408 pinned_vm:0x70c ret:0 [52.254084] perf_mmap user->locked_vm:0x408 pinned_vm:0x78d ret:0 [52.254263] perf_mmap user->locked_vm:0x408 pinned_vm:0x80e ret:0 The value of user->locked_vm increases to a limit then the memory is tracked by pinned_vm. During deallocation the size is subtracted from pinned_vm until it hits a limit. Then a larger value is subtracted from locked_vm leading to a large number (because of type unsigned): [64.267797] perf_mmap_close mmap_user->locked_vm:0x408 pinned_vm:0x78d [64.267826] perf_mmap_close mmap_user->locked_vm:0x408 pinned_vm:0x70c [64.267848] perf_mmap_close mmap_user->locked_vm:0x408 pinned_vm:0x68b [64.267869] perf_mmap_close mmap_user->locked_vm:0x408 pinned_vm:0x60a [64.267891] perf_mmap_close mmap_user->locked_vm:0x408 pinned_vm:0x589 [64.267911] perf_mmap_close mmap_user->locked_vm:0x408 pinned_vm:0x508 [64.267933] perf_mmap_close mmap_user->locked_vm:0x408 pinned_vm:0x487 [64.267952] perf_mmap_close mmap_user->locked_vm:0x408 pinned_vm:0x406 [64.268883] perf_mmap_close mmap_user->locked_vm:0x307 pinned_vm:0x406 [64.269117] perf_mmap_close mmap_user->locked_vm:0x206 pinned_vm:0x406 [64.269433] perf_mmap_close mmap_user->locked_vm:0x105 pinned_vm:0x406 [64.269536] perf_mmap_close mmap_user->locked_vm:0x4 pinned_vm:0x404 [64.269797] perf_mmap_close mmap_user->locked_vm:0xffffffffffffff84 pinned_vm:0x303 [64.270105] perf_mmap_close mmap_user->locked_vm:0xffffffffffffff04 pinned_vm:0x202 [64.270374] perf_mmap_close mmap_user->locked_vm:0xfffffffffffffe84 pinned_vm:0x101 [64.270628] perf_mmap_close mmap_user->locked_vm:0xfffffffffffffe04 pinned_vm:0x0 This value sticks for the user until system is rebooted, causing follow-on system calls using locked_vm resource limit to fail. Note: There is no issue using the normal trace buffer. In fact the issue is in perf_mmap_close(). During allocation auxiliary trace buffer memory is either traced as 'extra' and added to 'pinned_vm' or trace as 'user_extra' and added to 'locked_vm'. This applies for normal trace buffers and auxiliary trace buffer. However in function perf_mmap_close() all auxiliary trace buffer is subtraced from 'locked_vm' and never from 'pinned_vm'. This breaks the ballance. Signed-off-by: Thomas Richter Acked-by: Peter Zijlstra Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: acme@kernel.org Cc: gor@linux.ibm.com Cc: hechaol@fb.com Cc: heiko.carstens@de.ibm.com Cc: linux-perf-users@vger.kernel.org Cc: songliubraving@fb.com Fixes: d44248a41337 ("perf/core: Rework memory accounting in perf_mmap()") Link: https://lkml.kernel.org/r/20191021083354.67868-1-tmricht@linux.ibm.com [ Minor readability edits. ] Signed-off-by: Ingo Molnar --- kernel/events/core.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index 9ec0b0bfddbd..f5d7950d1931 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -5607,8 +5607,10 @@ static void perf_mmap_close(struct vm_area_struct *vma) perf_pmu_output_stop(event); /* now it's safe to free the pages */ - atomic_long_sub(rb->aux_nr_pages, &mmap_user->locked_vm); - atomic64_sub(rb->aux_mmap_locked, &vma->vm_mm->pinned_vm); + if (!rb->aux_mmap_locked) + atomic_long_sub(rb->aux_nr_pages, &mmap_user->locked_vm); + else + atomic64_sub(rb->aux_mmap_locked, &vma->vm_mm->pinned_vm); /* this has to be the last one */ rb_free_aux(rb);