self-profile: add LD_PRELOAD Support

This code is a simple implementation of my idea, with a focus on making
the self-profile portable.
It seems useful, even if there's a call to the main function, because
This method seems to reduce overhead compared to using "perf record" directly.
We can directly insert the code according to its original purpose.

Here are the test results from my Raspberry Pi 5, gcc version 12.2.0 (Debian 12.2.0-14)

$ uname -a
Linux paran 6.10.1-v8-16k+ ThinkOpenly#1 SMP PREEMPT Sat Jul 27 17:52:03 KST 2024 aarch64 GNU/Linux

$ make run
export PERF_COUNT_HW_CPU_CYCLES=1; ./test_profile
Sorting...
00: { "H", 107, 0.900000 }
01: { "I", 111, 0.900000 }
02: { "G", 117, 0.900000 }
03: { "E", 127, 0.900000 }
04: { "F", 147, 0.900000 }
05: { "A", 157, 0.900000 }
06: { "K", 157, 0.900000 }
07: { "L", 157, 0.900000 }
08: { "M", 157, 0.900000 }
09: { "N", 157, 0.900000 }
10: { "O", 157, 0.900000 }
11: { "P", 157, 0.900000 }
12: { "Z", 157, 0.900000 }
13: { "C", 175, 0.900000 }
14: { "J", 227, 0.900000 }
15: { "B", 517, 0.900000 }
16: { "D", 571, 0.900000 }
PERF_COUNT_HW_CPU_CYCLES(0): 7970
export PERF_COUNT_HW_CPU_CYCLES=1; LD_PRELOAD=self-profile.so ./preload_test_profile
Sorting...
PERF_COUNT_HW_CPU_CYCLES(0): 7444
export PERF_COUNT_HW_CPU_CYCLES=1; LD_PRELOAD=self-profile.so ./bsearch
Sorting...
PERF_COUNT_HW_CPU_CYCLES(0): 6779

Signed-off-by: Yunseong Kim <[email protected]>

Author: yskelg

Makefile

@@ -1,10 +1,22 @@
 CFLAGS=-O -Wno-unused-result
+
+all: test_profile self-profile preload_test_profile bsearch
+
 test_profile: test_profile.o
 
+self-profile:
+	$(CC) self-profile.c -o self-profile.so -fPIC -shared -ldl
+
+preload_test_profile: preload_test_profile.o
+
+bsearch: bsearch.o
+
 .PHONY: clean
 clean:
-	rm test_profile test_profile.o
+	rm test_profile test_profile.o self-profile.so preload_test_profile.o preload_test_profile bsearch.o bsearch
 
 .PHONY: run
 run: test_profile
 	export PERF_COUNT_HW_CPU_CYCLES=1; ./test_profile
+	export PERF_COUNT_HW_CPU_CYCLES=1; LD_PRELOAD=self-profile.so ./preload_test_profile
+	export PERF_COUNT_HW_CPU_CYCLES=1; LD_PRELOAD=self-profile.so ./bsearch

bsearch.c

@@ -0,0 +1,39 @@
+/*
+ * Reference:
+ *  - https://man7.org/linux/man-pages/man3/bsearch.3.html
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+struct mi {
+    int nr;
+    char *name;
+} months[] = {
+    { 1, "jan" }, { 2, "feb" }, { 3, "mar" }, { 4, "apr" },
+    { 5, "may" }, { 6, "jun" }, { 7, "jul" }, { 8, "aug" },
+    { 9, "sep" }, {10, "oct" }, {11, "nov" }, {12, "dec" }
+};
+#define nr_of_months (sizeof(months)/sizeof(months[0]))
+static int
+compmi(const void *m1, const void *m2)
+{
+    struct mi *mi1 = (struct mi *) m1;
+    struct mi *mi2 = (struct mi *) m2;
+    return strcmp(mi1->name, mi2->name);
+}
+int
+main(int argc, char **argv)
+{
+    int i;
+    qsort(months, nr_of_months, sizeof(struct mi), compmi);
+    for (i = 1; i < argc; i++) {
+        struct mi key, *res;
+        key.name = argv[i];
+        res = bsearch(&key, months, nr_of_months,
+                      sizeof(struct mi), compmi);
+        if (res == NULL)
+            return 1;
+    }
+    return 0;
+}

preload_test_profile.c

@@ -0,0 +1,44 @@
+#define _GNU_SOURCE
+#include <stdlib.h>
+
+/*
+       void qsort_r(void *base, size_t nmemb, size_t size,
+                  int (*compar)(const void *, const void *, void *),
+                  void *arg);
+ */
+
+struct item {
+    char *name;
+    int id;
+    double rating;
+} items[] = {
+    { "A", 157, 0.9 },
+    { "B", 517, 0.9 },
+    { "C", 175, 0.9 },
+    { "D", 571, 0.9 },
+    { "E", 127, 0.9 },
+    { "F", 147, 0.9 },
+    { "G", 117, 0.9 },
+    { "H", 107, 0.9 },
+    { "I", 111, 0.9 },
+    { "J", 227, 0.9 },
+    { "K", 157, 0.9 },
+    { "L", 157, 0.9 },
+    { "M", 157, 0.9 },
+    { "N", 157, 0.9 },
+    { "O", 157, 0.9 },
+    { "P", 157, 0.9 },
+    { "Z", 157, 0.9 }
+};
+
+int compare_items(const void *a, const void *b, void *arg) {
+    const struct item *itemL = a, *itemR = b;
+    if (itemL->id < itemR->id) return -1;
+    if (itemL->id > itemR->id) return 1;
+    return 0;
+}
+
+int main() {
+    qsort_r(items, sizeof(items)/sizeof(items[0]), sizeof(items[0]), compare_items, 0);
+    return 0;
+}

self-profile.c

@@ -0,0 +1,56 @@
+#define _GNU_SOURCE
+#include "self-profile.h"
+
+/*
+ * Reference:
+ *  Hook main() using LD_PRELOAD:
+ *  - https://gist.github.com/apsun/1e144bf7639b22ff0097171fa0f8c6b1
+ *  Using makefile, LD_PRELOAD to executable file:
+ *  - https://stackoverflow.com/questions/30594838/using-makefile-ld-preload-to-executable-file
+ */
+
+/* Trampoline for the real main() */
+static int (*main_orig)(int, char **, char **);
+
+/* main() that gets called by __libc_start_main() */
+int main_hook(int argc, char **argv, char **envp)
+{
+    PROFILE_BEGIN();
+    printf("Sorting...\n");
+    PROFILE_START();
+    int ret = main_orig(argc, argv, envp);
+    PROFILE_STOP();
+    /*
+     * I'm a bit concerned about this part of the code.
+     * It would be great if we could use the output section universally like main function.
+     * I'll think about it a bit more.
+     */
+    PROFILE_REPORT();
+    PROFILE_END();
+
+    return ret;
+}
+
+/*
+ * Wrapper for __libc_start_main() that replaces the real main
+ * function with our hooked version.
+ */
+int __libc_start_main(
+    int (*main)(int, char **, char **),
+    int argc,
+    char **argv,
+    int (*init)(int, char **, char **),
+    void (*fini)(void),
+    void (*rtld_fini)(void),
+    void *stack_end)
+{
+    /* Save the real main function address */
+    main_orig = main;
+
+    /* Find the real __libc_start_main()... */
+    typeof(&__libc_start_main) orig = dlsym(RTLD_NEXT, "__libc_start_main");
+
+    /* ... and call it with our custom main function */
+    return orig(main_hook, argc, argv, init, fini, rtld_fini, stack_end);
+}
+

self-profile.h

@@ -5,6 +5,7 @@
 #include <stdio.h>
 #include <string.h>
 #include <unistd.h>
+#include <dlfcn.h>
 #include <sys/ioctl.h>
 #include <linux/perf_event.h>
 #include <asm/unistd.h>