executor: add some code style checks

Move the test from pkg/csource to executor/
in order to be able to (1) run it on *.cc files,
(2) run on unprocessed *.h files, (3) produce line numbers.
Add a check for missed space after //.

6 files changed, 119 insertions, 12 deletions

diff --git a/executor/common.h b/executor/common.h
index a3f442edc..35e3c78da 100644
--- a/executor/common.h
+++ b/executor/common.h
@@ -9,7 +9,7 @@
 // - NORETURN/PRINTF/debug are removed
 // - exitf/fail are replaced with exit
 // - uintN types are replaced with uintN_t
-// - /*FOO*/ placeholders are replaced by actual values
+// - /*{{{FOO}}}*/ placeholders are replaced by actual values
 
 #ifndef _GNU_SOURCE
 #define _GNU_SOURCE
diff --git a/executor/common_bsd.h b/executor/common_bsd.h
index 17306fb96..2aa651dd3 100644
--- a/executor/common_bsd.h
+++ b/executor/common_bsd.h
@@ -61,7 +61,7 @@ static int inject_fault(int nth)
 
 	en.scope = FAULT_SCOPE_LWP;
 	en.mode = 0; // FAULT_MODE_NTH_ONESHOT
-	en.nth = nth + 2; //FAULT_NTH_MIN
+	en.nth = nth + 2; // FAULT_NTH_MIN
 	if (ioctl(fd, FAULT_IOC_ENABLE, &en) != 0)
 		fail("FAULT_IOC_ENABLE failed with nth=%d", nth);
 
diff --git a/executor/common_linux.h b/executor/common_linux.h
index 8c20101ec..b49e67542 100644
--- a/executor/common_linux.h
+++ b/executor/common_linux.h
@@ -2313,12 +2313,12 @@ error:
 #include <string.h>
 #include <sys/mount.h>
 
-//syz_mount_image(fs ptr[in, string[disk_filesystems]], dir ptr[in, filename], size intptr, nsegs len[segments], segments ptr[in, array[fs_image_segment]], flags flags[mount_flags], opts ptr[in, fs_options[vfat_options]])
-//fs_image_segment {
+// syz_mount_image(fs ptr[in, string[disk_filesystems]], dir ptr[in, filename], size intptr, nsegs len[segments], segments ptr[in, array[fs_image_segment]], flags flags[mount_flags], opts ptr[in, fs_options[vfat_options]])
+// fs_image_segment {
 //	data	ptr[in, array[int8]]
 //	size	len[data, intptr]
 //	offset	intptr
-//}
+// }
 static long syz_mount_image(volatile long fsarg, volatile long dir, volatile unsigned long size, volatile unsigned long nsegs, volatile long segments, volatile long flags, volatile long optsarg)
 {
 	char loopname[64], fs[32], opts[256];
diff --git a/executor/executor.cc b/executor/executor.cc
index 5b7371b1d..04d0259f6 100644
--- a/executor/executor.cc
+++ b/executor/executor.cc
@@ -362,7 +362,7 @@ int main(int argc, char** argv)
 	}
 	if (argc >= 2 && strcmp(argv[1], "setup_kcsan_filterlist") == 0) {
 #if SYZ_HAVE_KCSAN
-		setup_kcsan_filterlist(argv + 2, argc - 2, /*suppress=*/true);
+		setup_kcsan_filterlist(argv + 2, argc - 2, true);
 #else
 		fail("KCSAN is not implemented");
 #endif
diff --git a/executor/style_test.go b/executor/style_test.go
new file mode 100644
index 000000000..c3fde4889
--- /dev/null
+++ b/executor/style_test.go
@@ -0,0 +1,107 @@
+// Copyright 2020 syzkaller project authors. All rights reserved.
+// Use of this source code is governed by Apache 2 LICENSE that can be found in the LICENSE file.
+
+package executor
+
+import (
+	"bytes"
+	"io/ioutil"
+	"path/filepath"
+	"regexp"
+	"sort"
+	"strings"
+	"testing"
+)
+
+func TestExecutorMistakes(t *testing.T) {
+	checks := []*struct {
+		pattern     string
+		suppression string
+		message     string
+		tests       []string
+		commonOnly  bool
+	}{
+		{
+			pattern:    `\)\n\t*(debug|debug_dump_data)\(`,
+			message:    "debug() calls are stripped from C reproducers, this code will break. Use {} around debug() to fix",
+			commonOnly: true,
+			tests: []string{
+				`
+if (foo)
+	debug("foo failed");
+`, `
+	if (x + y)
+		debug_dump_data(data, len);
+`,
+			},
+		},
+		{
+			// These are also not properly stripped by pkg/csource.
+			pattern: `/\*[^{]`,
+			message: "Don't use /* */ block comments. Use // line comments instead",
+			tests: []string{
+				`/* C++ comment */`,
+			},
+		},
+		{
+			pattern:     `//[^\s]`,
+			suppression: `https?://`,
+			message:     "Add a space after //",
+			tests: []string{
+				`//foo`,
+			},
+		},
+	}
+	for _, check := range checks {
+		re := regexp.MustCompile(check.pattern)
+		for _, test := range check.tests {
+			if !re.MatchString(test) {
+				t.Fatalf("patter %q does not match test %q", check.pattern, test)
+			}
+		}
+	}
+	for _, file := range executorFiles(t) {
+		data, err := ioutil.ReadFile(file)
+		if err != nil {
+			t.Fatal(err)
+		}
+		for _, check := range checks {
+			if check.commonOnly && !strings.Contains(file, "common") {
+				continue
+			}
+			re := regexp.MustCompile(check.pattern)
+			supp := regexp.MustCompile(check.suppression)
+			for _, match := range re.FindAllIndex(data, -1) {
+				start, end := match[0], match[1]
+				for start != 0 && data[start-1] != '\n' {
+					start--
+				}
+				for end != len(data) && data[end] != '\n' {
+					end++
+				}
+				if supp.Match(data[start:end]) {
+					continue
+				}
+				line := bytes.Count(data[:start], []byte{'\n'}) + 1
+				t.Errorf("\nexecutor/%v:%v: %v\n%s\n", file, line, check.message, data[start:end])
+			}
+		}
+	}
+}
+
+func executorFiles(t *testing.T) []string {
+	cc, err := filepath.Glob("*.cc")
+	if err != nil {
+		t.Fatal(err)
+	}
+	h, err := filepath.Glob("*.h")
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(cc) == 0 || len(h) == 0 {
+		t.Fatal("found no executor files")
+	}
+	res := append(cc, h...)
+	sort.Strings(res)
+	return res
+}
diff --git a/executor/test_linux.h b/executor/test_linux.h
index cdeb72717..96e185175 100644
--- a/executor/test_linux.h
+++ b/executor/test_linux.h
@@ -96,13 +96,13 @@ static int test_kvm()
 	printf("host kernel version %u\n", ver);
 
 	// TODO: test VM mode.
-	//const char text16_vm[] = "\x48\xc7\xc3\xde\xc0\xad\x0b\x90\x90\x48\xc7\xc0\xef\xcd\xab\x00\xf4";
-	//if (res = test_one(64, text16_vm, sizeof(text16_vm) - 1, KVM_SETUP_VM, KVM_EXIT_HLT, true))
+	// const char text16_vm[] = "\x48\xc7\xc3\xde\xc0\xad\x0b\x90\x90\x48\xc7\xc0\xef\xcd\xab\x00\xf4";
+	// if (res = test_one(64, text16_vm, sizeof(text16_vm) - 1, KVM_SETUP_VM, KVM_EXIT_HLT, true))
 	//	return res;
 
-	/// TODO: test code executed in interrupt handlers.
-	//const char text32_div0[] = "\x31\xc0\xf7\xf0";
-	//if (res = test_one(32, text32_div0, sizeof(text32_div0)-1, 0, KVM_EXIT_HLT, true))
+	// TODO: test code executed in interrupt handlers.
+	// const char text32_div0[] = "\x31\xc0\xf7\xf0";
+	// if (res = test_one(32, text32_div0, sizeof(text32_div0)-1, 0, KVM_EXIT_HLT, true))
 	//	return res;
 
 	const char text8[] = "\x66\xb8\xde\xc0\xad\x0b";
@@ -147,7 +147,7 @@ static int test_kvm()
 		if ((res = test_one(8, text8_smm, sizeof(text8_smm) - 1, KVM_SETUP_SMM | KVM_SETUP_PROTECTED, KVM_EXIT_HLT, true)))
 			return res;
 
-		//const char text32_smm[] = "\xb8\xde\xc0\xad\x0b";
+		// const char text32_smm[] = "\xb8\xde\xc0\xad\x0b";
 		if ((res = test_one(32, text8_smm, sizeof(text8_smm) - 1, KVM_SETUP_SMM, KVM_EXIT_HLT, true)))
 			return res;