Skip to contentSkip to conversation

Miscellaneous fixes for BPF (perf tree)

PATCH Applied Mickaël Salaün 3 participants 10 messages Feb 7, 2017 – Feb 8, 2017

Subsystems (4)

v3 of 4 Applied Latest is v4 →

lore
Share & tools
lore
#0 lore raw

v3 cover Miscellaneous fixes for BPF (perf tree)

This series brings some fixes and small improvements to the BPF library and samples.

This is intended for the perf tree and apply on e06094ab6755 ("Merge remote-tracking branch 'tip/perf/urgent' into perf/core").

Changes since v2:
* add this cover letter

Changes since v1:
* exclude patches not intended for the perf tree

Regards,

Mickaël Salaün (5):
bpf: Add missing header to the library
bpf: Simplify bpf_load_program() error handling in the library samples/bpf: Ignore already processed ELF sections samples/bpf: Reset global variables
samples/bpf: Add missing header

 samples/bpf/bpf_load.c     |  7 +++++++
 samples/bpf/tracex5_kern.c |  1 +
 tools/lib/bpf/bpf.c        | 18 ++++++------------
 tools/lib/bpf/bpf.h        |  1 +
 4 files changed, 15 insertions(+), 12 deletions(-)
Signature · 1 line
2.11.0
#1 lore raw

PATCH v3 1/5 bpf: Add missing header to the library

Include stddef.h to define size_t.

Cc Alexei Starovoitov <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc Wang Nan <[email protected]>
---
 tools/lib/bpf/bpf.h | 1 +
 1 file changed, 1 insertion(+)
tools/lib/bpf/bpf.h +1 -0
index a2f9853dd882..df6e186da788 100644
@@ -22,6 +22,7 @@
 #define __BPF_BPF_H
 
 #include <linux/bpf.h>
+#include <stddef.h>
 
 int bpf_create_map(enum bpf_map_type map_type, int key_size, int value_size,
 		   int max_entries, __u32 map_flags);
Signature · 1 line
2.11.0
#2 lore raw

Re: [PATCH v3 1/5] bpf: Add missing header to the library

Quoted 16 lines · Mickaël Salaün

Include stddef.h to define size_t.

Cc Alexei Starovoitov <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc Wang Nan <[email protected]>
---
  tools/lib/bpf/bpf.h | 1 +
  1 file changed, 1 insertion(+)
tools/lib/bpf/bpf.h +0 -0
index a2f9853dd882..df6e186da788 100644

tools/lib/bpf/bpf.h line 28

Quoted 7 lines
tools/lib/bpf/bpf.h +1 -0
@@ -22,6 +22,7 @@
  #define __BPF_BPF_H
 
  #include <linux/bpf.h>
+#include <stddef.h>
 
  int bpf_create_map(enum bpf_map_type map_type, int key_size, int value_size,
  		   int max_entries, __u32 map_flags);

Looks good to me.

Thank you.

#3 lore raw

Re: [PATCH v3 1/5] bpf: Add missing header to the library

Em Wed, Feb 08, 2017 at 10:47:18AM +0800, Wangnan (F) escreveu:

Quoted 10 lines

tools/lib/bpf/bpf.h line 26

Quoted 7 lines

+++ b/tools/lib/bpf/bpf.h

tools/lib/bpf/bpf.h +1 -0
@@ -22,6 +22,7 @@
 #define __BPF_BPF_H
 #include <linux/bpf.h>
+#include <stddef.h>
 int bpf_create_map(enum bpf_map_type map_type, int key_size, int value_size,
 		   int max_entries, __u32 map_flags);

Looks good to me.

Thank you.

Applied, took the "Thank you" as an "Acked-by: Wang",

Regards,

- Arnaldo

#4 lore raw

PATCH v3 2/5 bpf: Simplify bpf_load_program() error handling in the library

In reply to Mickaël Salaün

Do not call a second time bpf(2) when a program load failed.

Cc Alexei Starovoitov <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc Wang Nan <[email protected]>
---
 tools/lib/bpf/bpf.c | 18 ++++++------------
 1 file changed, 6 insertions(+), 12 deletions(-)
tools/lib/bpf/bpf.c +6 -12
index 3ddb58a36d3c..fda3f494f1cd 100644
@@ -73,7 +73,6 @@ int bpf_load_program(enum bpf_prog_type type, struct bpf_insn *insns,
 		     size_t insns_cnt, char *license,
 		     __u32 kern_version, char *log_buf, size_t log_buf_sz)
 {
-	int fd;
 	union bpf_attr attr;
 
 	bzero(&attr, sizeof(attr));
@@ -81,20 +80,15 @@ int bpf_load_program(enum bpf_prog_type type, struct bpf_insn *insns,
 	attr.insn_cnt = (__u32)insns_cnt;
 	attr.insns = ptr_to_u64(insns);
 	attr.license = ptr_to_u64(license);
-	attr.log_buf = ptr_to_u64(NULL);
-	attr.log_size = 0;
-	attr.log_level = 0;
+	attr.log_buf = ptr_to_u64(log_buf);
+	attr.log_size = log_buf_sz;
 	attr.kern_version = kern_version;
 
-	fd = sys_bpf(BPF_PROG_LOAD, &attr, sizeof(attr));
-	if (fd >= 0 || !log_buf || !log_buf_sz)
-		return fd;
+	if (log_buf && log_buf_sz > 0) {
+		attr.log_level = 1;
+		log_buf[0] = 0;
+	}
 
-	/* Try again with log */
-	attr.log_buf = ptr_to_u64(log_buf);
-	attr.log_size = log_buf_sz;
-	attr.log_level = 1;
-	log_buf[0] = 0;
 	return sys_bpf(BPF_PROG_LOAD, &attr, sizeof(attr));
 }
 
Signature · 1 line
2.11.0
#5 lore raw

Re: [PATCH v3 2/5] bpf: Simplify bpf_load_program() error handling in the library

Quoted 1 line · Mickaël Salaün

Do not call a second time bpf(2) when a program load failed.

BPF_PROG_LOAD should success most of the time. Setting log_level to 0 by default and require log buffer when failure can make it faster in normal case.

Thank you.

Quoted 48 lines
Cc Alexei Starovoitov <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc Wang Nan <[email protected]>
---
  tools/lib/bpf/bpf.c | 18 ++++++------------
  1 file changed, 6 insertions(+), 12 deletions(-)
tools/lib/bpf/bpf.c +6 -12
index 3ddb58a36d3c..fda3f494f1cd 100644
@@ -73,7 +73,6 @@ int bpf_load_program(enum bpf_prog_type type, struct bpf_insn *insns,
  		     size_t insns_cnt, char *license,
  		     __u32 kern_version, char *log_buf, size_t log_buf_sz)
  {
-	int fd;
  	union bpf_attr attr;
 
  	bzero(&attr, sizeof(attr));
@@ -81,20 +80,15 @@ int bpf_load_program(enum bpf_prog_type type, struct bpf_insn *insns,
  	attr.insn_cnt = (__u32)insns_cnt;
  	attr.insns = ptr_to_u64(insns);
  	attr.license = ptr_to_u64(license);
-	attr.log_buf = ptr_to_u64(NULL);
-	attr.log_size = 0;
-	attr.log_level = 0;
+	attr.log_buf = ptr_to_u64(log_buf);
+	attr.log_size = log_buf_sz;
  	attr.kern_version = kern_version;
 
-	fd = sys_bpf(BPF_PROG_LOAD, &attr, sizeof(attr));
-	if (fd >= 0 || !log_buf || !log_buf_sz)
-		return fd;
+	if (log_buf && log_buf_sz > 0) {
+		attr.log_level = 1;
+		log_buf[0] = 0;
+	}
 
-	/* Try again with log */
-	attr.log_buf = ptr_to_u64(log_buf);
-	attr.log_size = log_buf_sz;
-	attr.log_level = 1;
-	log_buf[0] = 0;
  	return sys_bpf(BPF_PROG_LOAD, &attr, sizeof(attr));
  }
 
#6 lore raw

Re: [PATCH v3 2/5] bpf: Simplify bpf_load_program() error handling in the library

Quoted 8 lines · Wangnan (F)
Quoted 1 line · Mickaël Salaün

Do not call a second time bpf(2) when a program load failed.

BPF_PROG_LOAD should success most of the time. Setting log_level to 0 by default and require log buffer when failure can make it faster in normal case.

OK, I'm not sure if this is relevant for a test but I'm going to remove this patch.

Thanks

Quoted 53 lines

Thank you.

Quoted 47 lines
Cc Alexei Starovoitov <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc Wang Nan <[email protected]>
---
  tools/lib/bpf/bpf.c | 18 ++++++------------
  1 file changed, 6 insertions(+), 12 deletions(-)
tools/lib/bpf/bpf.c +6 -10
index 3ddb58a36d3c..fda3f494f1cd 100644
@@ -73,7 +73,6 @@ int bpf_load_program(enum bpf_prog_type type, struct
 bpf_insn *insns,
               size_t insns_cnt, char *license,
               __u32 kern_version, char *log_buf, size_t log_buf_sz)
  {
-    int fd;
      union bpf_attr attr;
        bzero(&attr, sizeof(attr));
@@ -81,20 +80,15 @@ int bpf_load_program(enum bpf_prog_type type,
 struct bpf_insn *insns,
      attr.insn_cnt = (__u32)insns_cnt;
      attr.insns = ptr_to_u64(insns);
      attr.license = ptr_to_u64(license);
-    attr.log_buf = ptr_to_u64(NULL);
-    attr.log_size = 0;
-    attr.log_level = 0;
+    attr.log_buf = ptr_to_u64(log_buf);
+    attr.log_size = log_buf_sz;
      attr.kern_version = kern_version;
  -    fd = sys_bpf(BPF_PROG_LOAD, &attr, sizeof(attr));
-    if (fd >= 0 || !log_buf || !log_buf_sz)
-        return fd;
+    if (log_buf && log_buf_sz > 0) {
+        attr.log_level = 1;
+        log_buf[0] = 0;
+    }
  -    /* Try again with log */
-    attr.log_buf = ptr_to_u64(log_buf);
-    attr.log_size = log_buf_sz;
-    attr.log_level = 1;
-    log_buf[0] = 0;
      return sys_bpf(BPF_PROG_LOAD, &attr, sizeof(attr));
  }
#7 lore raw

PATCH v3 3/5 samples/bpf: Ignore already processed ELF sections

In reply to Mickaël Salaün

Add a missing check for the map fixup loop.

Cc Alexei Starovoitov <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Daniel Borkmann <[email protected]>
---
 samples/bpf/bpf_load.c | 2 ++
 1 file changed, 2 insertions(+)
samples/bpf/bpf_load.c +2 -0
index 396e204888b3..e04fe09d7c2e 100644
@@ -328,6 +328,8 @@ int load_bpf_file(char *path)
 
 	/* load programs that need map fixup (relocations) */
 	for (i = 1; i < ehdr.e_shnum; i++) {
+		if (processed_sec[i])
+			continue;
 
 		if (get_sec(elf, i, &ehdr, &shname, &shdr, &data))
 			continue;
Signature · 1 line
2.11.0
#8 lore raw

PATCH v3 4/5 samples/bpf: Reset global variables

In reply to Mickaël Salaün

Before loading a new ELF, clean previous kernel version, license and processed sections.

Cc Alexei Starovoitov <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc David S. Miller <[email protected]>
---
 samples/bpf/bpf_load.c | 5 +++++
 1 file changed, 5 insertions(+)
samples/bpf/bpf_load.c +5 -0
index e04fe09d7c2e..b86ee54da2d1 100644
@@ -277,6 +277,11 @@ int load_bpf_file(char *path)
 	Elf_Data *data, *data_prog, *symbols = NULL;
 	char *shname, *shname_prog;
 
+	/* reset global variables */
+	kern_version = 0;
+	memset(license, 0, sizeof(license));
+	memset(processed_sec, 0, sizeof(processed_sec));
+
 	if (elf_version(EV_CURRENT) == EV_NONE)
 		return 1;
 
Signature · 1 line
2.11.0
#9 lore raw

PATCH v3 5/5 samples/bpf: Add missing header

In reply to Mickaël Salaün

Include unistd.h to define __NR_getuid and __NR_getsid.

Cc Alexei Starovoitov <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc David S. Miller <[email protected]>
---
 samples/bpf/tracex5_kern.c | 1 +
 1 file changed, 1 insertion(+)
samples/bpf/tracex5_kern.c +1 -0
index fd12d7154d42..7e4cf74553ff 100644
@@ -8,6 +8,7 @@
 #include <linux/version.h>
 #include <uapi/linux/bpf.h>
 #include <uapi/linux/seccomp.h>
+#include <uapi/linux/unistd.h>
 #include "bpf_helpers.h"
 
 #define PROG(F) SEC("kprobe/"__stringify(F)) int bpf_func_##F
Signature · 1 line
2.11.0