Add two tools: mkdoubledirs and gensamehash

mkdoubledirs creates pairs of directories a/a, b/b, c/c etc. This is to
test file lookup code: there must be a directory named blah only if the
current directory is named blah too.

gensamehash generates file names with the same XFS hash to test the
code that resolves collisions. In principle, can be updated for other
hashes: ext*, exfat, etc.

.gitignore

@@ -8,10 +8,12 @@ umka_fuse
 umka_os
 umka_ping
 umka_gen_devices_dat
-mkdirrange
+tools/mkdirrange
-mkfilepattern
+tools/mkdoubledirs
-lfbviewx
+tools/gensamehash
-randdir
+tools/mkfilepattern
+tools/lfbviewx
+tools/randdir
 *.img
 *.img.xz
 *.img.gz

img/makefile

@@ -1,5 +1,6 @@
 RANDDIR=../tools/randdir
 MKDIRRANGE=../tools/mkdirrange
+MKDOUBLEDIRS=../tools/mkdoubledirs
 MKFILEPATTERN=../tools/mkfilepattern
 DIRTOTEST=python3 ../tools/dirtotest.py
 MOUNT_OPT=-t xfs
@@ -15,7 +16,7 @@ s4k: xfs_v4_ftype0_s4k_b4k_n8k.img
 
 unicode: xfs_v4_unicode.img
 
-v5: xfs_v5_ftype1_s05k_b2k_n8k.img xfs_v5_files_s05k_b4k_n8k.img xfs_bigtime.img xfs_nrext64.img
+v5: xfs_v5_ftype1_s05k_b2k_n8k.img xfs_v5_files_s05k_b4k_n8k.img xfs_bigtime.img xfs_nrext64.img xfs_lookup_v5.img xfs_lookup_v4.img
 
 coverage: jfs.img xfs_borg_bit.img xfs_short_dir_i8.img
 
@@ -37,6 +38,76 @@ jfs.img:
 	parted --script $@ mktable gpt
 	parted --script --align optimal $@ mkpart primary 1MiB 100%
 
+xfs_lookup_v4.img:
+	fallocate -l 10GiB $@
+	mkfs.xfs -m crc=0 $@
+	sudo mount $(MOUNT_OPT) $@ $(TEMP_DIR)
+	sudo chown $$USER $(TEMP_DIR) -R
+#
+	mkdir $(TEMP_DIR)/dir_sf
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_sf d 5
+#
+	mkdir $(TEMP_DIR)/dir_block
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_block d 50
+#
+	mkdir $(TEMP_DIR)/dir_leaf
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_leaf d 500
+#
+	mkdir $(TEMP_DIR)/dir_node
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_node d 2000
+#
+	mkdir $(TEMP_DIR)/dir_btree_l1a
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_btree_l1a d 5000
+#
+	mkdir $(TEMP_DIR)/dir_btree_l1b
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_btree_l1b d 50000
+#
+	mkdir $(TEMP_DIR)/dir_btree_l1c
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_btree_l1c d 500000
+#
+	mkdir $(TEMP_DIR)/dir_btree_l2
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_btree_l2 d 2000000
+#
+	sudo umount $(TEMP_DIR)
+	fallocate -i -o 0 -l 1MiB $@
+	parted --script --align optimal $@ mktable msdos
+	parted --script --align optimal $@ mkpart primary xfs 1MiB 100%
+
+xfs_lookup_v5.img:
+	fallocate -l 10GiB $@
+	mkfs.xfs -m crc=1 $@
+	sudo mount $(MOUNT_OPT) $@ $(TEMP_DIR)
+	sudo chown $$USER $(TEMP_DIR) -R
+#
+	mkdir $(TEMP_DIR)/dir_sf
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_sf d 5
+#
+	mkdir $(TEMP_DIR)/dir_block
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_block d 50
+#
+	mkdir $(TEMP_DIR)/dir_leaf
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_leaf d 500
+#
+	mkdir $(TEMP_DIR)/dir_node
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_node d 2000
+#
+	mkdir $(TEMP_DIR)/dir_btree_l1a
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_btree_l1a d 5000
+#
+	mkdir $(TEMP_DIR)/dir_btree_l1b
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_btree_l1b d 50000
+#
+	mkdir $(TEMP_DIR)/dir_btree_l1c
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_btree_l1c d 500000
+#
+	mkdir $(TEMP_DIR)/dir_btree_l2
+	$(MKDOUBLEDIRS) $(TEMP_DIR)/dir_btree_l2 d 2000000
+#
+	sudo umount $(TEMP_DIR)
+	fallocate -i -o 0 -l 1MiB $@
+	parted --script --align optimal $@ mktable msdos
+	parted --script --align optimal $@ mkpart primary xfs 1MiB 100%
+
 xfs_nrext64.img:
 	fallocate -l 3000MiB $@
 	mkfs.xfs -i nrext64=1 $@
@@ -118,7 +189,7 @@ xfs_short_dir_i8.img:
 #
 	sudo umount $(TEMP_DIR)
 	fallocate -i -o 0 -l 1MiB $@
-	parted --script $@ mktable gpt
+	parted --script $@ mktable gpt	# FIXME: all msdos
 	parted --script --align optimal $@ mkpart primary 1MiB 100%
 
 xfs_v4_ftype0_s05k_b2k_n8k.img:

tools/gensamehash.c

@@ -0,0 +1,117 @@
+/*
+    SPDX-License-Identifier: GPL-2.0-or-later
+
+    UMKa - User-Mode KolibriOS developer tools
+    gensamehash - make directories with same names inside: dirname/dirname,
+    useful for search testing
+
+    Copyright (C) 2022  Ivan Baravy <dunkaist@gmail.com>
+*/
+
+#include <errno.h>
+#include <fcntl.h>
+#include <inttypes.h>
+#include <signal.h>
+#include <stdbool.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/stat.h>
+#include <sys/types.h>
+#include <unistd.h>
+
+#define NUM_LEN 18
+#define rol32(x,y)  (((x) << (y)) | ((x) >> (32 - (y))))
+
+// the hash function has been taken from XFS docs with minor edits
+uint32_t xfs_da_hashname(const char *name, int namelen) {
+    uint32_t hash;
+
+    /*
+     * Do four characters at a time as long as we can.
+     */
+    for (hash = 0; namelen >= 4; namelen -= 4, name += 4) {
+        hash = (name[0] << 21) ^ (name[1] << 14) ^ (name[2] << 7) ^
+               (name[3] << 0) ^ rol32(hash, 7 * 4);
+    }
+    /*
+     * Now do the rest of the characters.
+     */
+    switch (namelen) {
+    case 3:
+        return (name[0] << 14) ^ (name[1] << 7) ^ (name[2] << 0) ^
+               rol32(hash, 7 * 3);
+    case 2:
+        return (name[0] << 7) ^ (name[1] << 0) ^ rol32(hash, 7 * 2);
+    case 1:
+        return (name[0] << 0) ^ rol32(hash, 7 * 1);
+    default: /* case 0: */
+        return hash;
+    }
+}
+
+char name[256];
+char hash_filename[256];
+
+void increment(char *num) {
+    for (char *d = num + NUM_LEN; d > num; d--) {
+        if (*d < '9') {
+            d[0]++;
+            break;
+        } else {
+            *d = '0';
+        }
+    }
+}
+
+void int_handler(int signo, siginfo_t *info, void *context) {
+    (void)signo;
+    (void)info;
+    (void)context;
+    fprintf(stderr, "# cur name is %s\n", name);
+}
+
+int main(int argc, char *argv[])
+{
+    uint32_t hash;
+    int hash_cnt = 0;
+    uint64_t start_num;
+    if (argc != 4) {
+        fprintf(stderr, "gensamehash <hash> <hash_cnt> <start_num>\n");
+        exit(1);
+    }
+
+    hash = strtoul(argv[1], NULL, 16);
+    sscanf(argv[2], "%i", &hash_cnt);
+    sscanf(argv[3], "%" SCNu64, &start_num);
+
+    fprintf(stderr, "pid: %u\n", getpid());
+    sprintf(hash_filename, "hash_0x%8.8" PRIx32 ".%u", hash, getpid());
+
+    sprintf(name, "d%18.18" PRIu64, start_num);
+
+    struct sigaction sa;
+    sa.sa_sigaction = int_handler;
+    sigemptyset(&sa.sa_mask);
+    sa.sa_flags = SA_SIGINFO;
+
+    if (sigaction(SIGUSR1, &sa, NULL) == -1) {
+        fprintf(stderr, "Can't install SIGUSR1 handler!\n");
+        exit(1);
+    }
+
+    while (true) {
+        uint32_t h = xfs_da_hashname(name, NUM_LEN+1);
+        if (h == hash) {
+            FILE *f = fopen(hash_filename, "a+");
+            fprintf(f, "%7.7i %s\n", hash_cnt, name);
+            fclose(f);
+            hash_cnt++;
+        }
+//        fprintf(stderr, "# hash of %s is 0x%8.8x\n", name, h);
+        increment(name);
+    }
+
+    return 0;
+}

tools/makefile

@@ -7,7 +7,14 @@ CFLAGS=$(WARNINGS) $(NOWARNINGS) -std=c11 -O2 \
        -DNDEBUG -D_POSIX_C_SOURCE=200809L -fno-pie
 LDFLAGS=-no-pie
 
-all: mkdirrange mkfilepattern lfbviewx randdir covpreproc
+all: mkdirrange mkfilepattern lfbviewx randdir covpreproc mkdoubledirs \
+     gensamehash
+
+gensamehash: gensamehash.c
+	$(CC) $(CFLAGS) $(LDFLAGS) $< -o $@
+
+mkdoubledirs: mkdoubledirs.c
+	$(CC) $(CFLAGS) $(LDFLAGS) $< -o $@
 
 mkdirrange: mkdirrange.c
 	$(CC) $(CFLAGS) $(LDFLAGS) $< -o $@
@@ -27,4 +34,5 @@ covpreproc: covpreproc.c
 .PHONY: all clean
 
 clean:
-	rm -f *.o mkdirrange mkfilepattern lfbviewx randdir covpreproc
+	rm -f *.o mkdirrange mkfilepattern lfbviewx randdir covpreproc \
+          mkdoubledirs gensamehash

tools/mkdoubledirs.c

@@ -0,0 +1,58 @@
+/*
+    SPDX-License-Identifier: GPL-2.0-or-later
+
+    UMKa - User-Mode KolibriOS developer tools
+    mkdoubledirs - make directories with same names inside: dirname/dirname,
+    useful for search testing
+
+    Copyright (C) 2022  Ivan Baravy <dunkaist@gmail.com>
+*/
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <sys/stat.h>
+#include <sys/types.h>
+#include <fcntl.h>
+#include <errno.h>
+#include <string.h>
+#include <unistd.h>
+
+int main(int argc, char *argv[])
+{
+    const char *prefix;
+    const char *path;
+    unsigned count;
+    if (argc != 4) {
+        fprintf(stderr, "mkdoubledirs <directory> <prefix> <count>\n");
+        exit(1);
+    }
+    path = argv[1];
+    prefix = argv[2];
+    sscanf(argv[3], "%u", &count);
+
+    if (chdir(path)) {
+        fprintf(stderr, "Can't change dir to %s: %s\n", path, strerror(errno));
+        exit(1);
+    }
+    char dirname[256];
+
+    for(unsigned cur = 0; cur < count; cur++) {
+        sprintf(dirname, "%s%10.10u", prefix, cur);
+        if(mkdirat(AT_FDCWD, dirname, 0755)) {
+            fprintf(stderr, "Can't mkdir %s: %s\n", dirname, strerror(errno));
+            exit(1);
+        }
+        int dirfd = open(dirname, O_DIRECTORY);
+        if (dirfd == -1) {
+            fprintf(stderr, "Can't open %s: %s\n", dirname, strerror(errno));
+            exit(1);
+        }
+        if(mkdirat(dirfd, dirname, 0755)) {
+            fprintf(stderr, "Can't mkdir %s: %s\n", dirname, strerror(errno));
+            exit(1);
+        }
+        close(dirfd);
+    }
+
+    return 0;
+}