OpenCloudOS-Kernel/tools/perf/util/bpf-prologue.c

509 lines
12 KiB
C
Raw Normal View History

License cleanup: add SPDX GPL-2.0 license identifier to files with no license Many source files in the tree are missing licensing information, which makes it harder for compliance tools to determine the correct license. By default all files without license information are under the default license of the kernel, which is GPL version 2. Update the files which contain no license information with the 'GPL-2.0' SPDX license identifier. The SPDX identifier is a legally binding shorthand, which can be used instead of the full boiler plate text. This patch is based on work done by Thomas Gleixner and Kate Stewart and Philippe Ombredanne. How this work was done: Patches were generated and checked against linux-4.14-rc6 for a subset of the use cases: - file had no licensing information it it. - file was a */uapi/* one with no licensing information in it, - file was a */uapi/* one with existing licensing information, Further patches will be generated in subsequent months to fix up cases where non-standard license headers were used, and references to license had to be inferred by heuristics based on keywords. The analysis to determine which SPDX License Identifier to be applied to a file was done in a spreadsheet of side by side results from of the output of two independent scanners (ScanCode & Windriver) producing SPDX tag:value files created by Philippe Ombredanne. Philippe prepared the base worksheet, and did an initial spot review of a few 1000 files. The 4.13 kernel was the starting point of the analysis with 60,537 files assessed. Kate Stewart did a file by file comparison of the scanner results in the spreadsheet to determine which SPDX license identifier(s) to be applied to the file. She confirmed any determination that was not immediately clear with lawyers working with the Linux Foundation. Criteria used to select files for SPDX license identifier tagging was: - Files considered eligible had to be source code files. - Make and config files were included as candidates if they contained >5 lines of source - File already had some variant of a license header in it (even if <5 lines). All documentation files were explicitly excluded. The following heuristics were used to determine which SPDX license identifiers to apply. - when both scanners couldn't find any license traces, file was considered to have no license information in it, and the top level COPYING file license applied. For non */uapi/* files that summary was: SPDX license identifier # files ---------------------------------------------------|------- GPL-2.0 11139 and resulted in the first patch in this series. If that file was a */uapi/* path one, it was "GPL-2.0 WITH Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was: SPDX license identifier # files ---------------------------------------------------|------- GPL-2.0 WITH Linux-syscall-note 930 and resulted in the second patch in this series. - if a file had some form of licensing information in it, and was one of the */uapi/* ones, it was denoted with the Linux-syscall-note if any GPL family license was found in the file or had no licensing in it (per prior point). Results summary: SPDX license identifier # files ---------------------------------------------------|------ GPL-2.0 WITH Linux-syscall-note 270 GPL-2.0+ WITH Linux-syscall-note 169 ((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21 ((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17 LGPL-2.1+ WITH Linux-syscall-note 15 GPL-1.0+ WITH Linux-syscall-note 14 ((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5 LGPL-2.0+ WITH Linux-syscall-note 4 LGPL-2.1 WITH Linux-syscall-note 3 ((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3 ((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1 and that resulted in the third patch in this series. - when the two scanners agreed on the detected license(s), that became the concluded license(s). - when there was disagreement between the two scanners (one detected a license but the other didn't, or they both detected different licenses) a manual inspection of the file occurred. - In most cases a manual inspection of the information in the file resulted in a clear resolution of the license that should apply (and which scanner probably needed to revisit its heuristics). - When it was not immediately clear, the license identifier was confirmed with lawyers working with the Linux Foundation. - If there was any question as to the appropriate license identifier, the file was flagged for further research and to be revisited later in time. In total, over 70 hours of logged manual review was done on the spreadsheet to determine the SPDX license identifiers to apply to the source files by Kate, Philippe, Thomas and, in some cases, confirmation by lawyers working with the Linux Foundation. Kate also obtained a third independent scan of the 4.13 code base from FOSSology, and compared selected files where the other two scanners disagreed against that SPDX file, to see if there was new insights. The Windriver scanner is based on an older version of FOSSology in part, so they are related. Thomas did random spot checks in about 500 files from the spreadsheets for the uapi headers and agreed with SPDX license identifier in the files he inspected. For the non-uapi files Thomas did random spot checks in about 15000 files. In initial set of patches against 4.14-rc6, 3 files were found to have copy/paste license identifier errors, and have been fixed to reflect the correct identifier. Additionally Philippe spent 10 hours this week doing a detailed manual inspection and review of the 12,461 patched files from the initial patch version early this week with: - a full scancode scan run, collecting the matched texts, detected license ids and scores - reviewing anything where there was a license detected (about 500+ files) to ensure that the applied SPDX license was correct - reviewing anything where there was no detection but the patch license was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied SPDX license was correct This produced a worksheet with 20 files needing minor correction. This worksheet was then exported into 3 different .csv files for the different types of files to be modified. These .csv files were then reviewed by Greg. Thomas wrote a script to parse the csv files and add the proper SPDX tag to the file, in the format that the file expected. This script was further refined by Greg based on the output to detect more types of files automatically and to distinguish between header and source .c files (which need different comment types.) Finally Greg ran the script using the .csv files to generate the patches. Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org> Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 22:07:57 +08:00
// SPDX-License-Identifier: GPL-2.0
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
/*
* bpf-prologue.c
*
* Copyright (C) 2015 He Kuang <hekuang@huawei.com>
* Copyright (C) 2015 Wang Nan <wangnan0@huawei.com>
* Copyright (C) 2015 Huawei Inc.
*/
#include <bpf/libbpf.h>
#include "debug.h"
#include "bpf-loader.h"
#include "bpf-prologue.h"
#include "probe-finder.h"
#include <errno.h>
#include <stdlib.h>
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
#include <dwarf-regs.h>
#include <linux/filter.h>
#define BPF_REG_SIZE 8
#define JMP_TO_ERROR_CODE -1
#define JMP_TO_SUCCESS_CODE -2
#define JMP_TO_USER_CODE -3
struct bpf_insn_pos {
struct bpf_insn *begin;
struct bpf_insn *end;
struct bpf_insn *pos;
};
static inline int
pos_get_cnt(struct bpf_insn_pos *pos)
{
return pos->pos - pos->begin;
}
static int
append_insn(struct bpf_insn new_insn, struct bpf_insn_pos *pos)
{
if (!pos->pos)
return -BPF_LOADER_ERRNO__PROLOGUE2BIG;
if (pos->pos + 1 >= pos->end) {
pr_err("bpf prologue: prologue too long\n");
pos->pos = NULL;
return -BPF_LOADER_ERRNO__PROLOGUE2BIG;
}
*(pos->pos)++ = new_insn;
return 0;
}
static int
check_pos(struct bpf_insn_pos *pos)
{
if (!pos->pos || pos->pos >= pos->end)
return -BPF_LOADER_ERRNO__PROLOGUE2BIG;
return 0;
}
perf bpf: Fix endianness problem when loading parameters in prologue Perf's BPF prologue generator unconditionally fetches 8 bytes for function parameters, which causes problems on big endian machines. Thomas gives a detailed analysis for this problem: http://lkml.kernel.org/r/968ebda5-abe4-8830-8d69-49f62529d151@linux.vnet.ibm.com ---- 8< ---- I investigated perf test BPF for s390x and have a question regarding the 38.3 subtest (bpf-prologue test) which fails on s390x. When I turn on trace_printk in tests/bpf-script-test-prologue.c I see this output in /sys/kernel/debug/tracing/trace: [root@s8360047 perf]# cat /sys/kernel/debug/tracing/trace perf-30229 [000] d..2 170161.535791: : f_mode 2001d00000000 offset:0 orig:0 perf-30229 [000] d..2 170161.535809: : f_mode 6001f00000000 offset:0 orig:0 perf-30229 [000] d..2 170161.535815: : f_mode 6001f00000000 offset:1 orig:0 perf-30229 [000] d..2 170161.535819: : f_mode 2001d00000000 offset:1 orig:0 perf-30229 [000] d..2 170161.535822: : f_mode 2001d00000000 offset:2 orig:1 perf-30229 [000] d..2 170161.535825: : f_mode 6001f00000000 offset:2 orig:1 perf-30229 [000] d..2 170161.535828: : f_mode 6001f00000000 offset:3 orig:1 perf-30229 [000] d..2 170161.535832: : f_mode 2001d00000000 offset:3 orig:1 perf-30229 [000] d..2 170161.535835: : f_mode 2001d00000000 offset:4 orig:0 perf-30229 [000] d..2 170161.535841: : f_mode 6001f00000000 offset:4 orig:0 [...] There are 3 parameters the eBPF program tests/bpf-script-test-prologue.c accesses: f_mode (member of struct file at offset 140) offset and orig. They are parameters of the lseek() system call triggered in this test case in function llseek_loop(). What is really strange is the value of f_mode. It is an 8 byte value, whereas in the probe event it is defined as a 4 byte value. The lower 4 bytes are all zero and do not belong to member f_mode. The correct value should be 2001d for read-only and 6001f for read-write open mode. Here is the output of the 'perf test -vv bpf' trace: Try to find probe point from debuginfo. Matched function: null_lseek [2d9310d] Probe point found: null_lseek+0 Searching 'file' variable in context. Converting variable file into trace event. converting f_mode in file f_mode type is unsigned int. Opening /sys/kernel/debug/tracing//README write=0 Searching 'offset' variable in context. Converting variable offset into trace event. offset type is long long int. Searching 'orig' variable in context. Converting variable orig into trace event. orig type is int. Found 1 probe_trace_events. Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: p:perf_bpf_probe/func _text+8794224 f_mode=+140(%r2):x32 ---- 8< ---- This patch parses the type of each argument and converts data from memory to expected type. Now the test runs successfully on 4.13.0-rc5: [root@s8360046 perf]# ./perf test bpf 38: BPF filter : 38.1: Basic BPF filtering : Ok 38.2: BPF pinning : Ok 38.3: BPF prologue generation : Ok 38.4: BPF relocation checker : Ok [root@s8360046 perf]# Signed-off-by: Wang Nan <wangnan0@huawei.com> Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20170815092159.31912-1-tmricht@linux.vnet.ibm.com Signed-off-by: Thomas-Mich Richter <tmricht@linux.vnet.ibm.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-08-15 17:21:59 +08:00
/*
* Convert type string (u8/u16/u32/u64/s8/s16/s32/s64 ..., see
* Documentation/trace/kprobetrace.rst) to size field of BPF_LDX_MEM
perf bpf: Fix endianness problem when loading parameters in prologue Perf's BPF prologue generator unconditionally fetches 8 bytes for function parameters, which causes problems on big endian machines. Thomas gives a detailed analysis for this problem: http://lkml.kernel.org/r/968ebda5-abe4-8830-8d69-49f62529d151@linux.vnet.ibm.com ---- 8< ---- I investigated perf test BPF for s390x and have a question regarding the 38.3 subtest (bpf-prologue test) which fails on s390x. When I turn on trace_printk in tests/bpf-script-test-prologue.c I see this output in /sys/kernel/debug/tracing/trace: [root@s8360047 perf]# cat /sys/kernel/debug/tracing/trace perf-30229 [000] d..2 170161.535791: : f_mode 2001d00000000 offset:0 orig:0 perf-30229 [000] d..2 170161.535809: : f_mode 6001f00000000 offset:0 orig:0 perf-30229 [000] d..2 170161.535815: : f_mode 6001f00000000 offset:1 orig:0 perf-30229 [000] d..2 170161.535819: : f_mode 2001d00000000 offset:1 orig:0 perf-30229 [000] d..2 170161.535822: : f_mode 2001d00000000 offset:2 orig:1 perf-30229 [000] d..2 170161.535825: : f_mode 6001f00000000 offset:2 orig:1 perf-30229 [000] d..2 170161.535828: : f_mode 6001f00000000 offset:3 orig:1 perf-30229 [000] d..2 170161.535832: : f_mode 2001d00000000 offset:3 orig:1 perf-30229 [000] d..2 170161.535835: : f_mode 2001d00000000 offset:4 orig:0 perf-30229 [000] d..2 170161.535841: : f_mode 6001f00000000 offset:4 orig:0 [...] There are 3 parameters the eBPF program tests/bpf-script-test-prologue.c accesses: f_mode (member of struct file at offset 140) offset and orig. They are parameters of the lseek() system call triggered in this test case in function llseek_loop(). What is really strange is the value of f_mode. It is an 8 byte value, whereas in the probe event it is defined as a 4 byte value. The lower 4 bytes are all zero and do not belong to member f_mode. The correct value should be 2001d for read-only and 6001f for read-write open mode. Here is the output of the 'perf test -vv bpf' trace: Try to find probe point from debuginfo. Matched function: null_lseek [2d9310d] Probe point found: null_lseek+0 Searching 'file' variable in context. Converting variable file into trace event. converting f_mode in file f_mode type is unsigned int. Opening /sys/kernel/debug/tracing//README write=0 Searching 'offset' variable in context. Converting variable offset into trace event. offset type is long long int. Searching 'orig' variable in context. Converting variable orig into trace event. orig type is int. Found 1 probe_trace_events. Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: p:perf_bpf_probe/func _text+8794224 f_mode=+140(%r2):x32 ---- 8< ---- This patch parses the type of each argument and converts data from memory to expected type. Now the test runs successfully on 4.13.0-rc5: [root@s8360046 perf]# ./perf test bpf 38: BPF filter : 38.1: Basic BPF filtering : Ok 38.2: BPF pinning : Ok 38.3: BPF prologue generation : Ok 38.4: BPF relocation checker : Ok [root@s8360046 perf]# Signed-off-by: Wang Nan <wangnan0@huawei.com> Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20170815092159.31912-1-tmricht@linux.vnet.ibm.com Signed-off-by: Thomas-Mich Richter <tmricht@linux.vnet.ibm.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-08-15 17:21:59 +08:00
* instruction (BPF_{B,H,W,DW}).
*/
static int
argtype_to_ldx_size(const char *type)
{
int arg_size = type ? atoi(&type[1]) : 64;
switch (arg_size) {
case 8:
return BPF_B;
case 16:
return BPF_H;
case 32:
return BPF_W;
case 64:
default:
return BPF_DW;
}
}
static const char *
insn_sz_to_str(int insn_sz)
{
switch (insn_sz) {
case BPF_B:
return "BPF_B";
case BPF_H:
return "BPF_H";
case BPF_W:
return "BPF_W";
case BPF_DW:
return "BPF_DW";
default:
return "UNKNOWN";
}
}
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
/* Give it a shorter name */
#define ins(i, p) append_insn((i), (p))
/*
* Give a register name (in 'reg'), generate instruction to
* load register into an eBPF register rd:
* 'ldd target_reg, offset(ctx_reg)', where:
* ctx_reg is pre initialized to pointer of 'struct pt_regs'.
*/
static int
gen_ldx_reg_from_ctx(struct bpf_insn_pos *pos, int ctx_reg,
const char *reg, int target_reg)
{
int offset = regs_query_register_offset(reg);
if (offset < 0) {
pr_err("bpf: prologue: failed to get register %s\n",
reg);
return offset;
}
ins(BPF_LDX_MEM(BPF_DW, target_reg, ctx_reg, offset), pos);
return check_pos(pos);
}
/*
* Generate a BPF_FUNC_probe_read function call.
*
* src_base_addr_reg is a register holding base address,
* dst_addr_reg is a register holding dest address (on stack),
* result is:
*
* *[dst_addr_reg] = *([src_base_addr_reg] + offset)
*
* Arguments of BPF_FUNC_probe_read:
* ARG1: ptr to stack (dest)
* ARG2: size (8)
* ARG3: unsafe ptr (src)
*/
static int
gen_read_mem(struct bpf_insn_pos *pos,
int src_base_addr_reg,
int dst_addr_reg,
perf bpf: Fix bpf prologue generation Issue: bpf_probe_read() is no longer available for architecture which has overlapping address space. Hence bpf prologue generation fails Fix: Use bpf_probe_read_kernel for kernel member access. For user attribute access in kprobes, use bpf_probe_read_user. Other: @user attribute was introduced in commit 1e032f7cfa14 ("perf-probe: Add user memory access attribute support") Test: 1. ulimit -l 128 ; ./perf record -e tests/bpf_sched_setscheduler.c 2. cat tests/bpf_sched_setscheduler.c static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; 3. ./perf script sched 305669 [000] 1614458.838675: perf_bpf_probe:func: (2904e508) pid=261614 policy=2 sched_priority=1 4. cat /sys/kernel/debug/tracing/trace <...>-309956 [006] .... 1616098.093957: 0: prio: 1 Committer testing: I had to add some missing headers in the bpf_sched_setscheduler.c test proggie, then instead of using record+script I used 'perf trace' to drive everything in one go: # cat bpf_sched_setscheduler.c #include <linux/types.h> #include <bpf.h> static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; # # # perf trace -e bpf_sched_setscheduler.c chrt -f 42 sleep 1 0.000 chrt/80125 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) # And even with backtraces :-) # perf trace -e bpf_sched_setscheduler.c/max-stack=8/ chrt -f 42 sleep 1 0.000 chrt/79805 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) do_sched_setscheduler ([kernel.kallsyms]) __x64_sys_sched_setscheduler ([kernel.kallsyms]) do_syscall_64 ([kernel.kallsyms]) entry_SYSCALL_64 ([kernel.kallsyms]) __GI___sched_setscheduler (/usr/lib64/libc-2.30.so) # Signed-off-by: Sumanth Korikkar <sumanthk@linux.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Ilya Leoshkevich <iii@linux.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: bpf@vger.kernel.org LPU-Reference: 20200609081019.60234-3-sumanthk@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-06-09 16:10:19 +08:00
long offset,
int probeid)
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
{
/* mov arg3, src_base_addr_reg */
if (src_base_addr_reg != BPF_REG_ARG3)
ins(BPF_MOV64_REG(BPF_REG_ARG3, src_base_addr_reg), pos);
/* add arg3, #offset */
if (offset)
ins(BPF_ALU64_IMM(BPF_ADD, BPF_REG_ARG3, offset), pos);
/* mov arg2, #reg_size */
ins(BPF_ALU64_IMM(BPF_MOV, BPF_REG_ARG2, BPF_REG_SIZE), pos);
/* mov arg1, dst_addr_reg */
if (dst_addr_reg != BPF_REG_ARG1)
ins(BPF_MOV64_REG(BPF_REG_ARG1, dst_addr_reg), pos);
/* Call probe_read */
perf bpf: Fix bpf prologue generation Issue: bpf_probe_read() is no longer available for architecture which has overlapping address space. Hence bpf prologue generation fails Fix: Use bpf_probe_read_kernel for kernel member access. For user attribute access in kprobes, use bpf_probe_read_user. Other: @user attribute was introduced in commit 1e032f7cfa14 ("perf-probe: Add user memory access attribute support") Test: 1. ulimit -l 128 ; ./perf record -e tests/bpf_sched_setscheduler.c 2. cat tests/bpf_sched_setscheduler.c static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; 3. ./perf script sched 305669 [000] 1614458.838675: perf_bpf_probe:func: (2904e508) pid=261614 policy=2 sched_priority=1 4. cat /sys/kernel/debug/tracing/trace <...>-309956 [006] .... 1616098.093957: 0: prio: 1 Committer testing: I had to add some missing headers in the bpf_sched_setscheduler.c test proggie, then instead of using record+script I used 'perf trace' to drive everything in one go: # cat bpf_sched_setscheduler.c #include <linux/types.h> #include <bpf.h> static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; # # # perf trace -e bpf_sched_setscheduler.c chrt -f 42 sleep 1 0.000 chrt/80125 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) # And even with backtraces :-) # perf trace -e bpf_sched_setscheduler.c/max-stack=8/ chrt -f 42 sleep 1 0.000 chrt/79805 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) do_sched_setscheduler ([kernel.kallsyms]) __x64_sys_sched_setscheduler ([kernel.kallsyms]) do_syscall_64 ([kernel.kallsyms]) entry_SYSCALL_64 ([kernel.kallsyms]) __GI___sched_setscheduler (/usr/lib64/libc-2.30.so) # Signed-off-by: Sumanth Korikkar <sumanthk@linux.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Ilya Leoshkevich <iii@linux.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: bpf@vger.kernel.org LPU-Reference: 20200609081019.60234-3-sumanthk@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-06-09 16:10:19 +08:00
ins(BPF_EMIT_CALL(probeid), pos);
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
/*
* Error processing: if read fail, goto error code,
* will be relocated. Target should be the start of
* error processing code.
*/
ins(BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 0, JMP_TO_ERROR_CODE),
pos);
return check_pos(pos);
}
/*
* Each arg should be bare register. Fetch and save them into argument
* registers (r3 - r5).
*
* BPF_REG_1 should have been initialized with pointer to
* 'struct pt_regs'.
*/
static int
gen_prologue_fastpath(struct bpf_insn_pos *pos,
struct probe_trace_arg *args, int nargs)
{
int i, err = 0;
for (i = 0; i < nargs; i++) {
err = gen_ldx_reg_from_ctx(pos, BPF_REG_1, args[i].value,
BPF_PROLOGUE_START_ARG_REG + i);
if (err)
goto errout;
}
return check_pos(pos);
errout:
return err;
}
/*
* Slow path:
* At least one argument has the form of 'offset($rx)'.
*
* Following code first stores them into stack, then loads all of then
* to r2 - r5.
* Before final loading, the final result should be:
*
* low address
* BPF_REG_FP - 24 ARG3
* BPF_REG_FP - 16 ARG2
* BPF_REG_FP - 8 ARG1
* BPF_REG_FP
* high address
*
* For each argument (described as: offn(...off2(off1(reg)))),
* generates following code:
*
* r7 <- fp
* r7 <- r7 - stack_offset // Ideal code should initialize r7 using
* // fp before generating args. However,
* // eBPF won't regard r7 as stack pointer
* // if it is generated by minus 8 from
* // another stack pointer except fp.
* // This is why we have to set r7
* // to fp for each variable.
* r3 <- value of 'reg'-> generated using gen_ldx_reg_from_ctx()
* (r7) <- r3 // skip following instructions for bare reg
* r3 <- r3 + off1 . // skip if off1 == 0
* r2 <- 8 \
* r1 <- r7 |-> generated by gen_read_mem()
* call probe_read /
* jnei r0, 0, err ./
* r3 <- (r7)
* r3 <- r3 + off2 . // skip if off2 == 0
* r2 <- 8 \ // r2 may be broken by probe_read, so set again
* r1 <- r7 |-> generated by gen_read_mem()
* call probe_read /
* jnei r0, 0, err ./
* ...
*/
static int
gen_prologue_slowpath(struct bpf_insn_pos *pos,
struct probe_trace_arg *args, int nargs)
{
perf bpf: Fix bpf prologue generation Issue: bpf_probe_read() is no longer available for architecture which has overlapping address space. Hence bpf prologue generation fails Fix: Use bpf_probe_read_kernel for kernel member access. For user attribute access in kprobes, use bpf_probe_read_user. Other: @user attribute was introduced in commit 1e032f7cfa14 ("perf-probe: Add user memory access attribute support") Test: 1. ulimit -l 128 ; ./perf record -e tests/bpf_sched_setscheduler.c 2. cat tests/bpf_sched_setscheduler.c static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; 3. ./perf script sched 305669 [000] 1614458.838675: perf_bpf_probe:func: (2904e508) pid=261614 policy=2 sched_priority=1 4. cat /sys/kernel/debug/tracing/trace <...>-309956 [006] .... 1616098.093957: 0: prio: 1 Committer testing: I had to add some missing headers in the bpf_sched_setscheduler.c test proggie, then instead of using record+script I used 'perf trace' to drive everything in one go: # cat bpf_sched_setscheduler.c #include <linux/types.h> #include <bpf.h> static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; # # # perf trace -e bpf_sched_setscheduler.c chrt -f 42 sleep 1 0.000 chrt/80125 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) # And even with backtraces :-) # perf trace -e bpf_sched_setscheduler.c/max-stack=8/ chrt -f 42 sleep 1 0.000 chrt/79805 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) do_sched_setscheduler ([kernel.kallsyms]) __x64_sys_sched_setscheduler ([kernel.kallsyms]) do_syscall_64 ([kernel.kallsyms]) entry_SYSCALL_64 ([kernel.kallsyms]) __GI___sched_setscheduler (/usr/lib64/libc-2.30.so) # Signed-off-by: Sumanth Korikkar <sumanthk@linux.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Ilya Leoshkevich <iii@linux.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: bpf@vger.kernel.org LPU-Reference: 20200609081019.60234-3-sumanthk@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-06-09 16:10:19 +08:00
int err, i, probeid;
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
for (i = 0; i < nargs; i++) {
struct probe_trace_arg *arg = &args[i];
const char *reg = arg->value;
struct probe_trace_arg_ref *ref = NULL;
int stack_offset = (i + 1) * -8;
pr_debug("prologue: fetch arg %d, base reg is %s\n",
i, reg);
/* value of base register is stored into ARG3 */
err = gen_ldx_reg_from_ctx(pos, BPF_REG_CTX, reg,
BPF_REG_ARG3);
if (err) {
pr_err("prologue: failed to get offset of register %s\n",
reg);
goto errout;
}
/* Make r7 the stack pointer. */
ins(BPF_MOV64_REG(BPF_REG_7, BPF_REG_FP), pos);
/* r7 += -8 */
ins(BPF_ALU64_IMM(BPF_ADD, BPF_REG_7, stack_offset), pos);
/*
* Store r3 (base register) onto stack
* Ensure fp[offset] is set.
* fp is the only valid base register when storing
* into stack. We are not allowed to use r7 as base
* register here.
*/
ins(BPF_STX_MEM(BPF_DW, BPF_REG_FP, BPF_REG_ARG3,
stack_offset), pos);
ref = arg->ref;
perf bpf: Fix bpf prologue generation Issue: bpf_probe_read() is no longer available for architecture which has overlapping address space. Hence bpf prologue generation fails Fix: Use bpf_probe_read_kernel for kernel member access. For user attribute access in kprobes, use bpf_probe_read_user. Other: @user attribute was introduced in commit 1e032f7cfa14 ("perf-probe: Add user memory access attribute support") Test: 1. ulimit -l 128 ; ./perf record -e tests/bpf_sched_setscheduler.c 2. cat tests/bpf_sched_setscheduler.c static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; 3. ./perf script sched 305669 [000] 1614458.838675: perf_bpf_probe:func: (2904e508) pid=261614 policy=2 sched_priority=1 4. cat /sys/kernel/debug/tracing/trace <...>-309956 [006] .... 1616098.093957: 0: prio: 1 Committer testing: I had to add some missing headers in the bpf_sched_setscheduler.c test proggie, then instead of using record+script I used 'perf trace' to drive everything in one go: # cat bpf_sched_setscheduler.c #include <linux/types.h> #include <bpf.h> static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; # # # perf trace -e bpf_sched_setscheduler.c chrt -f 42 sleep 1 0.000 chrt/80125 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) # And even with backtraces :-) # perf trace -e bpf_sched_setscheduler.c/max-stack=8/ chrt -f 42 sleep 1 0.000 chrt/79805 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) do_sched_setscheduler ([kernel.kallsyms]) __x64_sys_sched_setscheduler ([kernel.kallsyms]) do_syscall_64 ([kernel.kallsyms]) entry_SYSCALL_64 ([kernel.kallsyms]) __GI___sched_setscheduler (/usr/lib64/libc-2.30.so) # Signed-off-by: Sumanth Korikkar <sumanthk@linux.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Ilya Leoshkevich <iii@linux.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: bpf@vger.kernel.org LPU-Reference: 20200609081019.60234-3-sumanthk@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-06-09 16:10:19 +08:00
probeid = BPF_FUNC_probe_read_kernel;
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
while (ref) {
pr_debug("prologue: arg %d: offset %ld\n",
i, ref->offset);
perf bpf: Fix bpf prologue generation Issue: bpf_probe_read() is no longer available for architecture which has overlapping address space. Hence bpf prologue generation fails Fix: Use bpf_probe_read_kernel for kernel member access. For user attribute access in kprobes, use bpf_probe_read_user. Other: @user attribute was introduced in commit 1e032f7cfa14 ("perf-probe: Add user memory access attribute support") Test: 1. ulimit -l 128 ; ./perf record -e tests/bpf_sched_setscheduler.c 2. cat tests/bpf_sched_setscheduler.c static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; 3. ./perf script sched 305669 [000] 1614458.838675: perf_bpf_probe:func: (2904e508) pid=261614 policy=2 sched_priority=1 4. cat /sys/kernel/debug/tracing/trace <...>-309956 [006] .... 1616098.093957: 0: prio: 1 Committer testing: I had to add some missing headers in the bpf_sched_setscheduler.c test proggie, then instead of using record+script I used 'perf trace' to drive everything in one go: # cat bpf_sched_setscheduler.c #include <linux/types.h> #include <bpf.h> static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; # # # perf trace -e bpf_sched_setscheduler.c chrt -f 42 sleep 1 0.000 chrt/80125 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) # And even with backtraces :-) # perf trace -e bpf_sched_setscheduler.c/max-stack=8/ chrt -f 42 sleep 1 0.000 chrt/79805 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) do_sched_setscheduler ([kernel.kallsyms]) __x64_sys_sched_setscheduler ([kernel.kallsyms]) do_syscall_64 ([kernel.kallsyms]) entry_SYSCALL_64 ([kernel.kallsyms]) __GI___sched_setscheduler (/usr/lib64/libc-2.30.so) # Signed-off-by: Sumanth Korikkar <sumanthk@linux.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Ilya Leoshkevich <iii@linux.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: bpf@vger.kernel.org LPU-Reference: 20200609081019.60234-3-sumanthk@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-06-09 16:10:19 +08:00
if (ref->user_access)
probeid = BPF_FUNC_probe_read_user;
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
err = gen_read_mem(pos, BPF_REG_3, BPF_REG_7,
perf bpf: Fix bpf prologue generation Issue: bpf_probe_read() is no longer available for architecture which has overlapping address space. Hence bpf prologue generation fails Fix: Use bpf_probe_read_kernel for kernel member access. For user attribute access in kprobes, use bpf_probe_read_user. Other: @user attribute was introduced in commit 1e032f7cfa14 ("perf-probe: Add user memory access attribute support") Test: 1. ulimit -l 128 ; ./perf record -e tests/bpf_sched_setscheduler.c 2. cat tests/bpf_sched_setscheduler.c static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; 3. ./perf script sched 305669 [000] 1614458.838675: perf_bpf_probe:func: (2904e508) pid=261614 policy=2 sched_priority=1 4. cat /sys/kernel/debug/tracing/trace <...>-309956 [006] .... 1616098.093957: 0: prio: 1 Committer testing: I had to add some missing headers in the bpf_sched_setscheduler.c test proggie, then instead of using record+script I used 'perf trace' to drive everything in one go: # cat bpf_sched_setscheduler.c #include <linux/types.h> #include <bpf.h> static void (*bpf_trace_printk)(const char *fmt, int fmt_size, ...) = (void *) 6; static int (*bpf_probe_read_user)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 112; static int (*bpf_probe_read_kernel)(void *dst, __u32 size, const void *unsafe_ptr) = (void *) 113; SEC("func=do_sched_setscheduler pid policy param->sched_priority@user") int bpf_func__setscheduler(void *ctx, int err, pid_t pid, int policy, int param) { char fmt[] = "prio: %ld"; bpf_trace_printk(fmt, sizeof(fmt), param); return 1; } char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; # # # perf trace -e bpf_sched_setscheduler.c chrt -f 42 sleep 1 0.000 chrt/80125 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) # And even with backtraces :-) # perf trace -e bpf_sched_setscheduler.c/max-stack=8/ chrt -f 42 sleep 1 0.000 chrt/79805 perf_bpf_probe:func(__probe_ip: -1676607808, policy: 1, sched_priority: 42) do_sched_setscheduler ([kernel.kallsyms]) __x64_sys_sched_setscheduler ([kernel.kallsyms]) do_syscall_64 ([kernel.kallsyms]) entry_SYSCALL_64 ([kernel.kallsyms]) __GI___sched_setscheduler (/usr/lib64/libc-2.30.so) # Signed-off-by: Sumanth Korikkar <sumanthk@linux.ibm.com> Reviewed-by: Thomas Richter <tmricht@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Ilya Leoshkevich <iii@linux.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: bpf@vger.kernel.org LPU-Reference: 20200609081019.60234-3-sumanthk@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-06-09 16:10:19 +08:00
ref->offset, probeid);
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
if (err) {
pr_err("prologue: failed to generate probe_read function call\n");
goto errout;
}
ref = ref->next;
/*
* Load previous result into ARG3. Use
* BPF_REG_FP instead of r7 because verifier
* allows FP based addressing only.
*/
if (ref)
ins(BPF_LDX_MEM(BPF_DW, BPF_REG_ARG3,
BPF_REG_FP, stack_offset), pos);
}
}
/* Final pass: read to registers */
perf bpf: Fix endianness problem when loading parameters in prologue Perf's BPF prologue generator unconditionally fetches 8 bytes for function parameters, which causes problems on big endian machines. Thomas gives a detailed analysis for this problem: http://lkml.kernel.org/r/968ebda5-abe4-8830-8d69-49f62529d151@linux.vnet.ibm.com ---- 8< ---- I investigated perf test BPF for s390x and have a question regarding the 38.3 subtest (bpf-prologue test) which fails on s390x. When I turn on trace_printk in tests/bpf-script-test-prologue.c I see this output in /sys/kernel/debug/tracing/trace: [root@s8360047 perf]# cat /sys/kernel/debug/tracing/trace perf-30229 [000] d..2 170161.535791: : f_mode 2001d00000000 offset:0 orig:0 perf-30229 [000] d..2 170161.535809: : f_mode 6001f00000000 offset:0 orig:0 perf-30229 [000] d..2 170161.535815: : f_mode 6001f00000000 offset:1 orig:0 perf-30229 [000] d..2 170161.535819: : f_mode 2001d00000000 offset:1 orig:0 perf-30229 [000] d..2 170161.535822: : f_mode 2001d00000000 offset:2 orig:1 perf-30229 [000] d..2 170161.535825: : f_mode 6001f00000000 offset:2 orig:1 perf-30229 [000] d..2 170161.535828: : f_mode 6001f00000000 offset:3 orig:1 perf-30229 [000] d..2 170161.535832: : f_mode 2001d00000000 offset:3 orig:1 perf-30229 [000] d..2 170161.535835: : f_mode 2001d00000000 offset:4 orig:0 perf-30229 [000] d..2 170161.535841: : f_mode 6001f00000000 offset:4 orig:0 [...] There are 3 parameters the eBPF program tests/bpf-script-test-prologue.c accesses: f_mode (member of struct file at offset 140) offset and orig. They are parameters of the lseek() system call triggered in this test case in function llseek_loop(). What is really strange is the value of f_mode. It is an 8 byte value, whereas in the probe event it is defined as a 4 byte value. The lower 4 bytes are all zero and do not belong to member f_mode. The correct value should be 2001d for read-only and 6001f for read-write open mode. Here is the output of the 'perf test -vv bpf' trace: Try to find probe point from debuginfo. Matched function: null_lseek [2d9310d] Probe point found: null_lseek+0 Searching 'file' variable in context. Converting variable file into trace event. converting f_mode in file f_mode type is unsigned int. Opening /sys/kernel/debug/tracing//README write=0 Searching 'offset' variable in context. Converting variable offset into trace event. offset type is long long int. Searching 'orig' variable in context. Converting variable orig into trace event. orig type is int. Found 1 probe_trace_events. Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: p:perf_bpf_probe/func _text+8794224 f_mode=+140(%r2):x32 ---- 8< ---- This patch parses the type of each argument and converts data from memory to expected type. Now the test runs successfully on 4.13.0-rc5: [root@s8360046 perf]# ./perf test bpf 38: BPF filter : 38.1: Basic BPF filtering : Ok 38.2: BPF pinning : Ok 38.3: BPF prologue generation : Ok 38.4: BPF relocation checker : Ok [root@s8360046 perf]# Signed-off-by: Wang Nan <wangnan0@huawei.com> Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20170815092159.31912-1-tmricht@linux.vnet.ibm.com Signed-off-by: Thomas-Mich Richter <tmricht@linux.vnet.ibm.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-08-15 17:21:59 +08:00
for (i = 0; i < nargs; i++) {
int insn_sz = (args[i].ref) ? argtype_to_ldx_size(args[i].type) : BPF_DW;
pr_debug("prologue: load arg %d, insn_sz is %s\n",
i, insn_sz_to_str(insn_sz));
ins(BPF_LDX_MEM(insn_sz, BPF_PROLOGUE_START_ARG_REG + i,
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
BPF_REG_FP, -BPF_REG_SIZE * (i + 1)), pos);
perf bpf: Fix endianness problem when loading parameters in prologue Perf's BPF prologue generator unconditionally fetches 8 bytes for function parameters, which causes problems on big endian machines. Thomas gives a detailed analysis for this problem: http://lkml.kernel.org/r/968ebda5-abe4-8830-8d69-49f62529d151@linux.vnet.ibm.com ---- 8< ---- I investigated perf test BPF for s390x and have a question regarding the 38.3 subtest (bpf-prologue test) which fails on s390x. When I turn on trace_printk in tests/bpf-script-test-prologue.c I see this output in /sys/kernel/debug/tracing/trace: [root@s8360047 perf]# cat /sys/kernel/debug/tracing/trace perf-30229 [000] d..2 170161.535791: : f_mode 2001d00000000 offset:0 orig:0 perf-30229 [000] d..2 170161.535809: : f_mode 6001f00000000 offset:0 orig:0 perf-30229 [000] d..2 170161.535815: : f_mode 6001f00000000 offset:1 orig:0 perf-30229 [000] d..2 170161.535819: : f_mode 2001d00000000 offset:1 orig:0 perf-30229 [000] d..2 170161.535822: : f_mode 2001d00000000 offset:2 orig:1 perf-30229 [000] d..2 170161.535825: : f_mode 6001f00000000 offset:2 orig:1 perf-30229 [000] d..2 170161.535828: : f_mode 6001f00000000 offset:3 orig:1 perf-30229 [000] d..2 170161.535832: : f_mode 2001d00000000 offset:3 orig:1 perf-30229 [000] d..2 170161.535835: : f_mode 2001d00000000 offset:4 orig:0 perf-30229 [000] d..2 170161.535841: : f_mode 6001f00000000 offset:4 orig:0 [...] There are 3 parameters the eBPF program tests/bpf-script-test-prologue.c accesses: f_mode (member of struct file at offset 140) offset and orig. They are parameters of the lseek() system call triggered in this test case in function llseek_loop(). What is really strange is the value of f_mode. It is an 8 byte value, whereas in the probe event it is defined as a 4 byte value. The lower 4 bytes are all zero and do not belong to member f_mode. The correct value should be 2001d for read-only and 6001f for read-write open mode. Here is the output of the 'perf test -vv bpf' trace: Try to find probe point from debuginfo. Matched function: null_lseek [2d9310d] Probe point found: null_lseek+0 Searching 'file' variable in context. Converting variable file into trace event. converting f_mode in file f_mode type is unsigned int. Opening /sys/kernel/debug/tracing//README write=0 Searching 'offset' variable in context. Converting variable offset into trace event. offset type is long long int. Searching 'orig' variable in context. Converting variable orig into trace event. orig type is int. Found 1 probe_trace_events. Opening /sys/kernel/debug/tracing//kprobe_events write=1 Writing event: p:perf_bpf_probe/func _text+8794224 f_mode=+140(%r2):x32 ---- 8< ---- This patch parses the type of each argument and converts data from memory to expected type. Now the test runs successfully on 4.13.0-rc5: [root@s8360046 perf]# ./perf test bpf 38: BPF filter : 38.1: Basic BPF filtering : Ok 38.2: BPF pinning : Ok 38.3: BPF prologue generation : Ok 38.4: BPF relocation checker : Ok [root@s8360046 perf]# Signed-off-by: Wang Nan <wangnan0@huawei.com> Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20170815092159.31912-1-tmricht@linux.vnet.ibm.com Signed-off-by: Thomas-Mich Richter <tmricht@linux.vnet.ibm.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-08-15 17:21:59 +08:00
}
perf bpf: Add prologue for BPF programs for fetching arguments This patch generates a prologue for a BPF program which fetches arguments for it. With this patch, the program can have arguments as follow: SEC("lock_page=__lock_page page->flags") int lock_page(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } This patch passes at most 3 arguments from r3, r4 and r5. r1 is still the ctx pointer. r2 is used to indicate if dereferencing was done successfully. This patch uses r6 to hold ctx (struct pt_regs) and r7 to hold stack pointer for result. Result of each arguments first store on stack: low address BPF_REG_FP - 24 ARG3 BPF_REG_FP - 16 ARG2 BPF_REG_FP - 8 ARG1 BPF_REG_FP high address Then loaded into r3, r4 and r5. The output prologue for offn(...off2(off1(reg)))) should be: r6 <- r1 // save ctx into a callee saved register r7 <- fp r7 <- r7 - stack_offset // pointer to result slot /* load r3 with the offset in pt_regs of 'reg' */ (r7) <- r3 // make slot valid r3 <- r3 + off1 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 // result put onto stack call probe_read // read unsafe pointer jnei r0, 0, err // error checking r3 <- (r7) // read result r3 <- r3 + off2 // prepare to read unsafe pointer r2 <- 8 r1 <- r7 call probe_read jnei r0, 0, err ... /* load r2, r3, r4 from stack */ goto success err: r2 <- 1 /* load r3, r4, r5 with 0 */ goto usercode success: r2 <- 0 usercode: r1 <- r6 // restore ctx // original user code If all of arguments reside in register (dereferencing is not required), gen_prologue_fastpath() will be used to create fast prologue: r3 <- (r1 + offset of reg1) r4 <- (r1 + offset of reg2) r5 <- (r1 + offset of reg3) r2 <- 0 P.S. eBPF calling convention is defined as: * r0 - return value from in-kernel function, and exit value for eBPF program * r1 - r5 - arguments from eBPF program to in-kernel function * r6 - r9 - callee saved registers that in-kernel function will preserve * r10 - read-only frame pointer to access stack Committer note: At least testing if it builds and loads: # cat test_probe_arg.c struct pt_regs; __attribute__((section("lock_page=__lock_page page->flags"), used)) int func(struct pt_regs *ctx, int err, unsigned long flags) { return 1; } char _license[] __attribute__((section("license"), used)) = "GPL"; int _version __attribute__((section("version"), used)) = 0x40300; # perf record -e ./test_probe_arg.c usleep 1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.016 MB perf.data ] # perf evlist perf_bpf_probe:lock_page # Signed-off-by: He Kuang <hekuang@huawei.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Wang Nan <wangnan0@huawei.com> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1447675815-166222-11-git-send-email-wangnan0@huawei.com Signed-off-by: Wang Nan <wangnan0@huawei.com> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-16 20:10:12 +08:00
ins(BPF_JMP_IMM(BPF_JA, BPF_REG_0, 0, JMP_TO_SUCCESS_CODE), pos);
return check_pos(pos);
errout:
return err;
}
static int
prologue_relocate(struct bpf_insn_pos *pos, struct bpf_insn *error_code,
struct bpf_insn *success_code, struct bpf_insn *user_code)
{
struct bpf_insn *insn;
if (check_pos(pos))
return -BPF_LOADER_ERRNO__PROLOGUE2BIG;
for (insn = pos->begin; insn < pos->pos; insn++) {
struct bpf_insn *target;
u8 class = BPF_CLASS(insn->code);
u8 opcode;
if (class != BPF_JMP)
continue;
opcode = BPF_OP(insn->code);
if (opcode == BPF_CALL)
continue;
switch (insn->off) {
case JMP_TO_ERROR_CODE:
target = error_code;
break;
case JMP_TO_SUCCESS_CODE:
target = success_code;
break;
case JMP_TO_USER_CODE:
target = user_code;
break;
default:
pr_err("bpf prologue: internal error: relocation failed\n");
return -BPF_LOADER_ERRNO__PROLOGUE;
}
insn->off = target - (insn + 1);
}
return 0;
}
int bpf__gen_prologue(struct probe_trace_arg *args, int nargs,
struct bpf_insn *new_prog, size_t *new_cnt,
size_t cnt_space)
{
struct bpf_insn *success_code = NULL;
struct bpf_insn *error_code = NULL;
struct bpf_insn *user_code = NULL;
struct bpf_insn_pos pos;
bool fastpath = true;
int err = 0, i;
if (!new_prog || !new_cnt)
return -EINVAL;
if (cnt_space > BPF_MAXINSNS)
cnt_space = BPF_MAXINSNS;
pos.begin = new_prog;
pos.end = new_prog + cnt_space;
pos.pos = new_prog;
if (!nargs) {
ins(BPF_ALU64_IMM(BPF_MOV, BPF_PROLOGUE_FETCH_RESULT_REG, 0),
&pos);
if (check_pos(&pos))
goto errout;
*new_cnt = pos_get_cnt(&pos);
return 0;
}
if (nargs > BPF_PROLOGUE_MAX_ARGS) {
pr_warning("bpf: prologue: %d arguments are dropped\n",
nargs - BPF_PROLOGUE_MAX_ARGS);
nargs = BPF_PROLOGUE_MAX_ARGS;
}
/* First pass: validation */
for (i = 0; i < nargs; i++) {
struct probe_trace_arg_ref *ref = args[i].ref;
if (args[i].value[0] == '@') {
/* TODO: fetch global variable */
pr_err("bpf: prologue: global %s%+ld not support\n",
args[i].value, ref ? ref->offset : 0);
return -ENOTSUP;
}
while (ref) {
/* fastpath is true if all args has ref == NULL */
fastpath = false;
/*
* Instruction encodes immediate value using
* s32, ref->offset is long. On systems which
* can't fill long in s32, refuse to process if
* ref->offset too large (or small).
*/
#ifdef __LP64__
#define OFFSET_MAX ((1LL << 31) - 1)
#define OFFSET_MIN ((1LL << 31) * -1)
if (ref->offset > OFFSET_MAX ||
ref->offset < OFFSET_MIN) {
pr_err("bpf: prologue: offset out of bound: %ld\n",
ref->offset);
return -BPF_LOADER_ERRNO__PROLOGUEOOB;
}
#endif
ref = ref->next;
}
}
pr_debug("prologue: pass validation\n");
if (fastpath) {
/* If all variables are registers... */
pr_debug("prologue: fast path\n");
err = gen_prologue_fastpath(&pos, args, nargs);
if (err)
goto errout;
} else {
pr_debug("prologue: slow path\n");
/* Initialization: move ctx to a callee saved register. */
ins(BPF_MOV64_REG(BPF_REG_CTX, BPF_REG_ARG1), &pos);
err = gen_prologue_slowpath(&pos, args, nargs);
if (err)
goto errout;
/*
* start of ERROR_CODE (only slow pass needs error code)
* mov r2 <- 1 // r2 is error number
* mov r3 <- 0 // r3, r4... should be touched or
* // verifier would complain
* mov r4 <- 0
* ...
* goto usercode
*/
error_code = pos.pos;
ins(BPF_ALU64_IMM(BPF_MOV, BPF_PROLOGUE_FETCH_RESULT_REG, 1),
&pos);
for (i = 0; i < nargs; i++)
ins(BPF_ALU64_IMM(BPF_MOV,
BPF_PROLOGUE_START_ARG_REG + i,
0),
&pos);
ins(BPF_JMP_IMM(BPF_JA, BPF_REG_0, 0, JMP_TO_USER_CODE),
&pos);
}
/*
* start of SUCCESS_CODE:
* mov r2 <- 0
* goto usercode // skip
*/
success_code = pos.pos;
ins(BPF_ALU64_IMM(BPF_MOV, BPF_PROLOGUE_FETCH_RESULT_REG, 0), &pos);
/*
* start of USER_CODE:
* Restore ctx to r1
*/
user_code = pos.pos;
if (!fastpath) {
/*
* Only slow path needs restoring of ctx. In fast path,
* register are loaded directly from r1.
*/
ins(BPF_MOV64_REG(BPF_REG_ARG1, BPF_REG_CTX), &pos);
err = prologue_relocate(&pos, error_code, success_code,
user_code);
if (err)
goto errout;
}
err = check_pos(&pos);
if (err)
goto errout;
*new_cnt = pos_get_cnt(&pos);
return 0;
errout:
return err;
}