mirrors/cosmopolitan
1
0
Fork 0
mirror of https://github.com/jart/cosmopolitan.git synced 2025-02-01 03:53:33 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions
cosmopolitan/test/libc/intrin/strcmp_test.c

656 lines
29 KiB
C
Raw Normal View History

Initial import
2020-06-15 14:18:57 +00:00
/*-*- mode:c;indent-tabs-mode:nil;c-basic-offset:2;tab-width:8;coding:utf-8 -*-│
│vi: set net ft=c ts=2 sts=2 sw=2 fenc=utf-8 :vi│
╞══════════════════════════════════════════════════════════════════════════════╡
│ Copyright 2020 Justine Alexandra Roberts Tunney │
│ │
Change license
2020-12-28 01:18:44 +00:00
│ Permission to use, copy, modify, and/or distribute this software for │
│ any purpose with or without fee is hereby granted, provided that the │
│ above copyright notice and this permission notice appear in all copies. │
Initial import
2020-06-15 14:18:57 +00:00
│ │
Change license
2020-12-28 01:18:44 +00:00
│ THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL │
│ WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED │
│ WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE │
│ AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL │
│ DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR │
│ PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER │
│ TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR │
│ PERFORMANCE OF THIS SOFTWARE. │
Initial import
2020-06-15 14:18:57 +00:00
╚─────────────────────────────────────────────────────────────────────────────*/
Add SNI support to redbean and improve SSL perf This change makes SSL virtual hosting possible. You can now load multiple certificates for multiple domains and redbean will just figure out which one to use, even if you only have 1 ip address. You can also use a jumbo certificate that lists all your domains in the the subject alternative names. This change also makes performance improvements to MbedTLS. Here are some benchmarks vs. cc1920749eb81f346badaf55fbf79620cb718a55 BEFORE AFTER (microsecs) suite_ssl.com 2512881 191738 13.11x faster suite_pkparse.com 36291 3295 11.01x faster suite_x509parse.com 854669 120293 7.10x faster suite_pkwrite.com 6549 1265 5.18x faster suite_ecdsa.com 53347 18778 2.84x faster suite_pk.com 49051 18717 2.62x faster suite_ecdh.com 19535 9502 2.06x faster suite_shax.com 15848 7965 1.99x faster suite_rsa.com 353257 184828 1.91x faster suite_x509write.com 162646 85733 1.90x faster suite_ecp.com 20503 11050 1.86x faster suite_hmac_drbg.no_reseed.com 19528 11417 1.71x faster suite_hmac_drbg.nopr.com 12460 8010 1.56x faster suite_mpi.com 687124 442661 1.55x faster suite_hmac_drbg.pr.com 11890 7752 1.53x faster There aren't any special tricks to the performance imporvements. It's mostly due to code cleanup, assembly and intel instructions like mulx, adox, and adcx.
2021-07-19 21:55:20 +00:00
#include "libc/assert.h"
Initial import
2020-06-15 14:18:57 +00:00
#include "libc/dce.h"
Reduce build graph by another eight percent
2022-08-13 20:11:56 +00:00
#include "libc/intrin/bits.h"
Remove more nonstandard stuff from cosmopolitan.h Fixes #61
2021-03-01 07:42:35 +00:00
#include "libc/macros.internal.h"
Rename rand64() to _rand64()
2022-10-10 11:12:06 +00:00
#include "libc/mem/gc.internal.h"
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
#include "libc/mem/mem.h"
Initial import
2020-06-15 14:18:57 +00:00
#include "libc/nexgen32e/cachesize.h"
#include "libc/nexgen32e/x86feature.h"
Fix some build errors
2023-06-05 22:50:15 +00:00
#include "libc/runtime/runtime.h"
Reduce build graph by another eight percent
2022-08-13 20:11:56 +00:00
#include "libc/stdio/rand.h"
Initial import
2020-06-15 14:18:57 +00:00
#include "libc/stdio/stdio.h"
#include "libc/str/str.h"
#include "libc/testlib/ezbench.h"
Add SNI support to redbean and improve SSL perf This change makes SSL virtual hosting possible. You can now load multiple certificates for multiple domains and redbean will just figure out which one to use, even if you only have 1 ip address. You can also use a jumbo certificate that lists all your domains in the the subject alternative names. This change also makes performance improvements to MbedTLS. Here are some benchmarks vs. cc1920749eb81f346badaf55fbf79620cb718a55 BEFORE AFTER (microsecs) suite_ssl.com 2512881 191738 13.11x faster suite_pkparse.com 36291 3295 11.01x faster suite_x509parse.com 854669 120293 7.10x faster suite_pkwrite.com 6549 1265 5.18x faster suite_ecdsa.com 53347 18778 2.84x faster suite_pk.com 49051 18717 2.62x faster suite_ecdh.com 19535 9502 2.06x faster suite_shax.com 15848 7965 1.99x faster suite_rsa.com 353257 184828 1.91x faster suite_x509write.com 162646 85733 1.90x faster suite_ecp.com 20503 11050 1.86x faster suite_hmac_drbg.no_reseed.com 19528 11417 1.71x faster suite_hmac_drbg.nopr.com 12460 8010 1.56x faster suite_mpi.com 687124 442661 1.55x faster suite_hmac_drbg.pr.com 11890 7752 1.53x faster There aren't any special tricks to the performance imporvements. It's mostly due to code cleanup, assembly and intel instructions like mulx, adox, and adcx.
2021-07-19 21:55:20 +00:00
#include "libc/testlib/hyperion.h"
Initial import
2020-06-15 14:18:57 +00:00
#include "libc/testlib/testlib.h"
Make improvements
2020-12-01 11:43:40 +00:00
int (*memcmpi)(const void *, const void *, size_t) = memcmp;
Initial import
2020-06-15 14:18:57 +00:00
/*───────────────────────────────────────────────────────────────────────────│─╗
│ test/libc/str/strcmp_test.c § emptiness ─╬─│┼
╚────────────────────────────────────────────────────────────────────────────│*/
TEST(strcmp, emptyString) {
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_EQ(0, strcmp("", ""));
EXPECT_NE(0, strcmp("", "a"));
Initial import
2020-06-15 14:18:57 +00:00
}
TEST(strcasecmp, emptyString) {
EXPECT_EQ(0, strcasecmp("", ""));
EXPECT_NE(0, strcasecmp("", "a"));
}
TEST(strcmp16, emptyString) {
EXPECT_EQ(0, strcmp16(u"", u""));
EXPECT_NE(0, strcmp16(u"", u"a"));
}
TEST(strcasecmp16, emptyString) {
EXPECT_EQ(0, strcasecmp16(u"", u""));
EXPECT_NE(0, strcasecmp16(u"", u"a"));
}
TEST(wcscmp, emptyString) {
EXPECT_EQ(0, wcscmp(L"", L""));
EXPECT_NE(0, wcscmp(L"", L"a"));
}
TEST(wcscasecmp, emptyString) {
EXPECT_EQ(0, wcscasecmp(L"", L""));
EXPECT_NE(0, wcscasecmp(L"", L"a"));
}
Get more Python tests passing (#141)
2021-08-16 22:26:31 +00:00
TEST(strncmp, nullString) {
char *s1 = malloc(0);
char *s2 = malloc(0);
ASSERT_NE(s1, s2);
ASSERT_EQ(0, strncmp(s1, s2, 0));
free(s2);
free(s1);
}
Initial import
2020-06-15 14:18:57 +00:00
TEST(strncmp, emptyString) {
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
char *s1 = strcpy(malloc(1), "");
char *s2 = strcpy(malloc(1), "");
Make improvements
2020-12-01 11:43:40 +00:00
ASSERT_EQ(0, strncmp(s1, s2, 0));
ASSERT_EQ(0, strncmp(s1, s2, 1));
ASSERT_EQ(0, strncmp(s1, s2, -1));
ASSERT_EQ(0, strncmp(s1, s1, -1));
ASSERT_EQ(0, strncmp(s2, s2, -1));
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(s2);
free(s1);
Initial import
2020-06-15 14:18:57 +00:00
}
TEST(strncasecmp, emptyString) {
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
char *s1 = strcpy(malloc(1), "");
char *s2 = strcpy(malloc(1), "");
Initial import
2020-06-15 14:18:57 +00:00
ASSERT_EQ(0, strncasecmp(s1, s2, 0));
ASSERT_EQ(0, strncasecmp(s1, s2, 1));
ASSERT_EQ(0, strncasecmp(s1, s2, -1));
ASSERT_EQ(0, strncasecmp(s1, s1, -1));
ASSERT_EQ(0, strncasecmp(s2, s2, -1));
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(s2);
free(s1);
Initial import
2020-06-15 14:18:57 +00:00
}
/*───────────────────────────────────────────────────────────────────────────│─╗
│ test/libc/str/strcmp_test.c § inequality ─╬─│┼
╚────────────────────────────────────────────────────────────────────────────│*/
TEST(strncmp, testInequality) {
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
char *s1 = strcpy(malloc(2), "1");
char *s2 = strcpy(malloc(1), "");
Make improvements
2020-12-01 11:43:40 +00:00
ASSERT_EQ(0, strncmp(s1, s2, 0));
Make improvements This change progresses our AARCH64 support: - The AARCH64 build and tests are now passing - Add 128-bit floating-point support to printf() - Fix clone() so it initializes cosmo's x28 TLS register - Fix TLS memory layout issue with aarch64 _Alignas vars - Revamp microbenchmarking tools so they work on aarch64 - Make some subtle improvements to aarch64 crash reporting - Make kisdangerous() memory checks more accurate on aarch64 - Remove sys_open() since it's not available on Linux AARCH64 This change makes general improvements to Cosmo and Redbean: - Introduce GetHostIsa() function in Redbean - You can now feature check using pledge(0, 0) - You can now feature check using unveil("",0) - Refactor some more x86-specific asm comments - Refactor and write docs for some libm functions - Make the mmap() API behave more similar to Linux - Fix WIFSIGNALED() which wrongly returned true for zero - Rename some obscure cosmo keywords from noFOO to dontFOO
2023-06-03 15:12:13 +00:00
ASSERT_GT(strncmp(s1, s2, 1), 0);
ASSERT_LT(strncmp(s2, s1, 1), 0);
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(s2);
free(s1);
Initial import
2020-06-15 14:18:57 +00:00
}
/*───────────────────────────────────────────────────────────────────────────│─╗
│ test/libc/str/strcmp_test.c § does it work? ─╬─│┼
╚────────────────────────────────────────────────────────────────────────────│*/
TEST(memcmp, test) {
EXPECT_EQ(memcmpi("\200", "\200", 1), 0);
EXPECT_LT(memcmpi("\177", "\200", 1), 0);
EXPECT_GT(memcmpi("\200", "\177", 1), 0);
EXPECT_EQ(memcmpi("", "", 0), 0);
EXPECT_EQ(memcmpi("a", "a", 1), 0);
EXPECT_GT(memcmpi("a", "A", 1), 0);
EXPECT_LT(memcmpi("A", "a", 1), 0);
EXPECT_LT(memcmpi("\001", "\377", 1), 0);
EXPECT_GT(memcmpi("\377", "\001", 1), 0);
EXPECT_EQ(memcmpi("a", "aa", 1), 0);
EXPECT_EQ(memcmpi("aa", "a", 1), 0);
EXPECT_LT(memcmpi("a", "aa", 2), 0);
EXPECT_GT(memcmpi("aa", "a", 2), 0);
EXPECT_LT(memcmpi("aaaaaaaaaaaaaaa\001", "aaaaaaaaaaaaaaa\377", 16), 0);
EXPECT_GT(memcmpi("aaaaaaaaaaaaaaa\377", "aaaaaaaaaaaaaaa\001", 16), 0);
EXPECT_LT(memcmpi("aaaaaaaaaaaaaaaa\001", "aaaaaaaaaaaaaaaa\377", 17), 0);
EXPECT_GT(memcmpi("aaaaaaaaaaaaaaaa\377", "aaaaaaaaaaaaaaaa\001", 17), 0);
EXPECT_LT(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377", 32),
0);
EXPECT_GT(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001", 32),
0);
EXPECT_LT(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377", 33),
0);
EXPECT_GT(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001", 33),
0);
EXPECT_LT(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377", 34),
0);
EXPECT_GT(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001", 34),
0);
}
TEST(strcmp, testItWorks) {
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_EQ(strcmp("", ""), 0);
EXPECT_EQ(strcmp("a", "a"), 0);
EXPECT_GT(strcmp("a", "A"), 0);
EXPECT_LT(strcmp("A", "a"), 0);
EXPECT_LT(strcmp("\001", "\377"), 0);
EXPECT_GT(strcmp("\377", "\001"), 0);
EXPECT_LT(strcmp("a", "aa"), 0);
EXPECT_GT(strcmp("aa", "a"), 0);
EXPECT_LT(strcmp("a\000", "aa\000"), 0);
EXPECT_GT(strcmp("aa\000", "a\000"), 0);
EXPECT_LT(strcmp("aaaaaaaaaaaaaaa\001", "aaaaaaaaaaaaaaa\377"), 0);
EXPECT_GT(strcmp("aaaaaaaaaaaaaaa\377", "aaaaaaaaaaaaaaa\001"), 0);
EXPECT_LT(strcmp("aaaaaaaaaaaaaaaa\001", "aaaaaaaaaaaaaaaa\377"), 0);
EXPECT_GT(strcmp("aaaaaaaaaaaaaaaa\377", "aaaaaaaaaaaaaaaa\001"), 0);
EXPECT_LT(strcmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377"),
Initial import
2020-06-15 14:18:57 +00:00
0);
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_GT(strcmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001"),
Initial import
2020-06-15 14:18:57 +00:00
0);
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_LT(strcmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377"),
Initial import
2020-06-15 14:18:57 +00:00
0);
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_GT(strcmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001"),
Initial import
2020-06-15 14:18:57 +00:00
0);
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_LT(strcmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377"),
Initial import
2020-06-15 14:18:57 +00:00
0);
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_GT(strcmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001"),
Initial import
2020-06-15 14:18:57 +00:00
0);
}
TEST(strcasecmp, testItWorks) {
EXPECT_EQ(strcasecmp("", ""), 0);
EXPECT_EQ(strcasecmp("a", "a"), 0);
EXPECT_EQ(strcasecmp("a", "A"), 0);
EXPECT_EQ(strcasecmp("A", "a"), 0);
EXPECT_LT(strcasecmp("a", "z"), 0);
EXPECT_GT(strcasecmp("z", "a"), 0);
EXPECT_LT(strcasecmp("\001", "\377"), 0);
EXPECT_GT(strcasecmp("\377", "\001"), 0);
EXPECT_LT(strcasecmp("a", "aa"), 0);
EXPECT_GT(strcasecmp("aa", "a"), 0);
EXPECT_LT(strcasecmp("a\000", "aa\000"), 0);
EXPECT_GT(strcasecmp("aa\000", "a\000"), 0);
EXPECT_LT(strcasecmp("aaaaaaaaaaaaaaa\001", "aaaaaaaaaaaaaaa\377"), 0);
EXPECT_GT(strcasecmp("aaaaaaaaaaaaaaa\377", "aaaaaaaaaaaaaaa\001"), 0);
EXPECT_LT(strcasecmp("aaaaaaaaaaaaaaaa\001", "aaaaaaaaaaaaaaaa\377"), 0);
EXPECT_GT(strcasecmp("aaaaaaaaaaaaaaaa\377", "aaaaaaaaaaaaaaaa\001"), 0);
EXPECT_LT(strcasecmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(strcasecmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001"),
0);
EXPECT_LT(strcasecmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(strcasecmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001"),
0);
EXPECT_LT(strcasecmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(strcasecmp("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001"),
0);
}
TEST(strcmp16, testItWorks) {
EXPECT_EQ(strcmp16(u"", u""), 0);
EXPECT_EQ(strcmp16(u"a", u"a"), 0);
EXPECT_GT(strcmp16(u"a", u"A"), 0);
EXPECT_LT(strcmp16(u"A", u"a"), 0);
EXPECT_LT(strcmp16(u"\001", u"\377"), 0);
EXPECT_GT(strcmp16(u"\377", u"\001"), 0);
EXPECT_LT(strcmp16(u"a", u"aa"), 0);
EXPECT_GT(strcmp16(u"aa", u"a"), 0);
EXPECT_LT(strcmp16(u"a\000", u"aa\000"), 0);
EXPECT_GT(strcmp16(u"aa\000", u"a\000"), 0);
EXPECT_LT(strcmp16(u"aaaaaaaaaaaaaaa\001", u"aaaaaaaaaaaaaaa\377"), 0);
EXPECT_GT(strcmp16(u"aaaaaaaaaaaaaaa\377", u"aaaaaaaaaaaaaaa\001"), 0);
EXPECT_LT(strcmp16(u"aaaaaaaaaaaaaaaa\001", u"aaaaaaaaaaaaaaaa\377"), 0);
EXPECT_GT(strcmp16(u"aaaaaaaaaaaaaaaa\377", u"aaaaaaaaaaaaaaaa\001"), 0);
EXPECT_LT(strcmp16(u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001",
u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(strcmp16(u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377",
u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001"),
0);
EXPECT_LT(strcmp16(u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001",
u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(strcmp16(u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377",
u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001"),
0);
EXPECT_LT(strcmp16(u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001",
u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(strcmp16(u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377",
u"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001"),
0);
}
TEST(wcscmp, testItWorks) {
EXPECT_EQ(wcscmp(L"", L""), 0);
EXPECT_EQ(wcscmp(L"a", L"a"), 0);
EXPECT_GT(wcscmp(L"a", L"A"), 0);
EXPECT_LT(wcscmp(L"A", L"a"), 0);
EXPECT_LT(wcscmp(L"\001", L"\377"), 0);
EXPECT_GT(wcscmp(L"\377", L"\001"), 0);
EXPECT_LT(wcscmp(L"a", L"aa"), 0);
EXPECT_GT(wcscmp(L"aa", L"a"), 0);
EXPECT_LT(wcscmp(L"a", L"aa"), 0);
EXPECT_GT(wcscmp(L"aa", L"a"), 0);
EXPECT_LT(wcscmp(L"aaaaaaaaaaaaaaa\001", L"aaaaaaaaaaaaaaa\377"), 0);
EXPECT_GT(wcscmp(L"aaaaaaaaaaaaaaa\377", L"aaaaaaaaaaaaaaa\001"), 0);
EXPECT_LT(wcscmp(L"aaaaaaaaaaaaaaaa\001", L"aaaaaaaaaaaaaaaa\377"), 0);
EXPECT_GT(wcscmp(L"aaaaaaaaaaaaaaaa\377", L"aaaaaaaaaaaaaaaa\001"), 0);
EXPECT_LT(wcscmp(L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001",
L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(wcscmp(L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\377",
L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaa\001"),
0);
EXPECT_LT(wcscmp(L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001",
L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(wcscmp(L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\377",
L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaa\001"),
0);
EXPECT_LT(wcscmp(L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001",
L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377"),
0);
EXPECT_GT(wcscmp(L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\377",
L"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\001"),
0);
}
TEST(strcasecmp, testItWorksCase) {
EXPECT_EQ(0, strcasecmp("hello", "HELLO"));
EXPECT_EQ(0, strcasecmp("hello", "Hello"));
EXPECT_EQ(0, strcasecmp("hello", "hello"));
EXPECT_NE(0, strcasecmp("hello", "yello"));
}
TEST(strcasecmp16, testItWorksCase) {
EXPECT_EQ(0, strcasecmp16(u"hello", u"HELLO"));
EXPECT_EQ(0, strcasecmp16(u"hello", u"Hello"));
EXPECT_EQ(0, strcasecmp16(u"hello", u"hello"));
EXPECT_NE(0, strcasecmp16(u"hello", u"yello"));
}
TEST(wcscasecmp, testItWorksCase) {
EXPECT_EQ(0, wcscasecmp(L"hello", L"HELLO"));
EXPECT_EQ(0, wcscasecmp(L"hello", L"Hello"));
EXPECT_EQ(0, wcscasecmp(L"hello", L"hello"));
EXPECT_NE(0, wcscasecmp(L"hello", L"yello"));
}
/*───────────────────────────────────────────────────────────────────────────│─╗
│ test/libc/str/strcmp_test.c § nontrivial length ─╬─│┼
╚────────────────────────────────────────────────────────────────────────────│*/
TEST(strncmp, testEqualManyNs) {
Make more fixes and improvements - Remove PAGESIZE constant - Fix realloc() documentation - Fix ttyname_r() error reporting - Make forking more reliable on Windows - Make execvp() a few microseconds faster - Make system() a few microseconds faster - Tighten up the socket-related magic numbers - Loosen restrictions on mmap() offset alignment - Improve GetProgramExecutableName() with getenv("_") - Use mkstemp() as basis for mktemp(), tmpfile(), tmpfd() - Fix flakes in pthread_cancel_test, unix_test, fork_test - Fix recently introduced futex stack overflow regression - Let sockets be passed as stdio to subprocesses on Windows - Improve security of bind() on Windows w/ SO_EXCLUSIVEADDRUSE
2023-07-30 01:44:15 +00:00
char *s1 = malloc(4096);
char *s2 = malloc(4096);
memset(s1, 7, 4096);
memset(s2, 7, 4096);
s1[4096 - 1] = '\0';
s2[4096 - 1] = '\0';
Initial import
2020-06-15 14:18:57 +00:00
for (unsigned i = 1; i <= 128; ++i) {
Make more fixes and improvements - Remove PAGESIZE constant - Fix realloc() documentation - Fix ttyname_r() error reporting - Make forking more reliable on Windows - Make execvp() a few microseconds faster - Make system() a few microseconds faster - Tighten up the socket-related magic numbers - Loosen restrictions on mmap() offset alignment - Improve GetProgramExecutableName() with getenv("_") - Use mkstemp() as basis for mktemp(), tmpfile(), tmpfd() - Fix flakes in pthread_cancel_test, unix_test, fork_test - Fix recently introduced futex stack overflow regression - Let sockets be passed as stdio to subprocesses on Windows - Improve security of bind() on Windows w/ SO_EXCLUSIVEADDRUSE
2023-07-30 01:44:15 +00:00
ASSERT_EQ(0, strncmp(s1 + 4096 - i, s2 + 4096 - i, i + 0));
ASSERT_EQ(0, strncmp(s1 + 4096 - i, s2 + 4096 - i, i + 1));
Initial import
2020-06-15 14:18:57 +00:00
}
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(s2);
free(s1);
Initial import
2020-06-15 14:18:57 +00:00
}
TEST(strncmp, testNotEqualManyNs) {
Make more fixes and improvements - Remove PAGESIZE constant - Fix realloc() documentation - Fix ttyname_r() error reporting - Make forking more reliable on Windows - Make execvp() a few microseconds faster - Make system() a few microseconds faster - Tighten up the socket-related magic numbers - Loosen restrictions on mmap() offset alignment - Improve GetProgramExecutableName() with getenv("_") - Use mkstemp() as basis for mktemp(), tmpfile(), tmpfd() - Fix flakes in pthread_cancel_test, unix_test, fork_test - Fix recently introduced futex stack overflow regression - Let sockets be passed as stdio to subprocesses on Windows - Improve security of bind() on Windows w/ SO_EXCLUSIVEADDRUSE
2023-07-30 01:44:15 +00:00
char *s1 = malloc(4096);
char *s2 = malloc(4096);
Initial import
2020-06-15 14:18:57 +00:00
for (unsigned i = 1; i <= 128; ++i) {
Make more fixes and improvements - Remove PAGESIZE constant - Fix realloc() documentation - Fix ttyname_r() error reporting - Make forking more reliable on Windows - Make execvp() a few microseconds faster - Make system() a few microseconds faster - Tighten up the socket-related magic numbers - Loosen restrictions on mmap() offset alignment - Improve GetProgramExecutableName() with getenv("_") - Use mkstemp() as basis for mktemp(), tmpfile(), tmpfd() - Fix flakes in pthread_cancel_test, unix_test, fork_test - Fix recently introduced futex stack overflow regression - Let sockets be passed as stdio to subprocesses on Windows - Improve security of bind() on Windows w/ SO_EXCLUSIVEADDRUSE
2023-07-30 01:44:15 +00:00
memset(s1, 7, 4096);
memset(s2, 7, 4096);
s1[4096 - 1] = (unsigned char)0;
s2[4096 - 1] = (unsigned char)255;
ASSERT_EQ(-255, strncmp(s1 + 4096 - i, s2 + 4096 - i, i + 0));
ASSERT_EQ(-255, strncmp(s1 + 4096 - i, s2 + 4096 - i, i + 1));
Initial import
2020-06-15 14:18:57 +00:00
}
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(s2);
free(s1);
Initial import
2020-06-15 14:18:57 +00:00
}
/*───────────────────────────────────────────────────────────────────────────│─╗
│ test/libc/str/strcmp_test.c § nul termination vs. explicit length ─╬─│┼
╚────────────────────────────────────────────────────────────────────────────│*/
TEST(strncmp, testStringNulTerminatesBeforeExplicitLength) {
const char kRdi[] = "";
const char kRsi[] = "TZ=America/Los_Angeles";
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
char *rdi = memcpy(malloc(sizeof(kRdi)), kRdi, sizeof(kRdi));
char *rsi = memcpy(malloc(sizeof(kRsi)), kRsi, sizeof(kRsi));
Initial import
2020-06-15 14:18:57 +00:00
size_t rdx = 3;
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_EQ(strncmp(rdi, rdi, rdx), 0);
EXPECT_LT(strncmp(rdi, rsi, rdx), 0);
EXPECT_GT(strncmp(rsi, rdi, rdx), 0);
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(rsi);
free(rdi);
Initial import
2020-06-15 14:18:57 +00:00
}
TEST(strncasecmp, testStringNulTerminatesBeforeExplicitLength) {
const char kRdi[] = "";
const char kRsi[] = "TZ=America/Los_Angeles";
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
char *rdi = memcpy(malloc(sizeof(kRdi)), kRdi, sizeof(kRdi));
char *rsi = memcpy(malloc(sizeof(kRsi)), kRsi, sizeof(kRsi));
Initial import
2020-06-15 14:18:57 +00:00
size_t rdx = 3;
EXPECT_EQ(strncasecmp(rdi, rdi, rdx), 0);
EXPECT_LT(strncasecmp(rdi, rsi, rdx), 0);
EXPECT_GT(strncasecmp(rsi, rdi, rdx), 0);
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(rsi);
free(rdi);
Initial import
2020-06-15 14:18:57 +00:00
}
TEST(strncmp16, testStringNulTerminatesBeforeExplicitLength) {
const char16_t kRdi[] = u"";
const char16_t kRsi[] = u"TZ=America/Los_Angeles";
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
char16_t *rdi = memcpy(malloc(sizeof(kRdi)), kRdi, sizeof(kRdi));
char16_t *rsi = memcpy(malloc(sizeof(kRsi)), kRsi, sizeof(kRsi));
Initial import
2020-06-15 14:18:57 +00:00
size_t rdx = 3;
EXPECT_EQ(strncmp16(rdi, rdi, rdx), 0);
EXPECT_LT(strncmp16(rdi, rsi, rdx), 0);
EXPECT_GT(strncmp16(rsi, rdi, rdx), 0);
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(rsi);
free(rdi);
Initial import
2020-06-15 14:18:57 +00:00
}
/*───────────────────────────────────────────────────────────────────────────│─╗
│ test/libc/str/strcmp_test.c § two's complement bane ─╬─│┼
╚────────────────────────────────────────────────────────────────────────────│*/
TEST(strcmp, testTwosComplementBane_hasUnsignedBehavior) {
Make improvements
2020-12-01 11:43:40 +00:00
EXPECT_EQ(strcmp("\200", "\200"), 0);
EXPECT_LT(strcmp("\x7f", "\x80"), 0);
EXPECT_GT(strcmp("\x80", "\x7f"), 0);
Initial import
2020-06-15 14:18:57 +00:00
}
TEST(strcasecmp, testTwosComplementBane_hasUnsignedBehavior) {
EXPECT_EQ(strcasecmp("\200", "\200"), 0);
EXPECT_LT(strcasecmp("\x7f", "\x80"), 0);
EXPECT_GT(strcasecmp("\x80", "\x7f"), 0);
}
TEST(memcmp, testTwosComplementBane_unsignedBehavior) {
EXPECT_EQ(memcmpi("\200", "\200", 1), 0);
EXPECT_LT(memcmpi("\177", "\200", 1), 0);
EXPECT_GT(memcmpi("\200", "\177", 1), 0);
EXPECT_EQ(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\200",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\200", 34),
0);
EXPECT_LT(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\177",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\200", 34),
0);
EXPECT_GT(memcmpi("aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\200",
"aaaaaaaaabaaaaaaaaaaaaaaaaaaaaaaa\177", 34),
0);
}
TEST(strcmp16, testTwosComplementBane_hasUnsignedBehavior) {
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
char16_t *B1 = malloc(8);
char16_t *B2 = malloc(8);
Initial import
2020-06-15 14:18:57 +00:00
B1[1] = L'\0';
B2[1] = L'\0';
EXPECT_EQ(strcmp16(memcpy(B1, "\x00\x80", 2), memcpy(B2, "\x00\x80", 2)), 0);
EXPECT_LT(strcmp16(memcpy(B1, "\xff\x7f", 2), memcpy(B2, "\x00\x80", 2)), 0);
EXPECT_GT(strcmp16(memcpy(B1, "\x00\x80", 2), memcpy(B2, "\xff\x7f", 2)), 0);
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(B2);
free(B1);
Initial import
2020-06-15 14:18:57 +00:00
}
TEST(strncmp16, testTwosComplementBane_hasUnsignedBehavior) {
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
char16_t *B1 = malloc(4);
char16_t *B2 = malloc(4);
Initial import
2020-06-15 14:18:57 +00:00
EXPECT_EQ(strncmp16(memcpy(B1, "\x00\x80", 2), memcpy(B2, "\x00\x80", 2), 1),
0);
EXPECT_LT(strncmp16(memcpy(B1, "\xff\x7f", 2), memcpy(B2, "\x00\x80", 2), 1),
0);
EXPECT_GT(strncmp16(memcpy(B1, "\x00\x80", 2), memcpy(B2, "\xff\x7f", 2), 1),
0);
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(B2);
free(B1);
Initial import
2020-06-15 14:18:57 +00:00
}
Make improvements
2020-12-01 11:43:40 +00:00
TEST(wcscmp, testTwosComplementBane) {
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
wchar_t *B1 = malloc(8);
wchar_t *B2 = malloc(8);
Initial import
2020-06-15 14:18:57 +00:00
B1[1] = L'\0';
B2[1] = L'\0';
EXPECT_EQ(wcscmp(memcpy(B1, "\x00\x00\x00\x80", 4),
memcpy(B2, "\x00\x00\x00\x80", 4)),
0);
Fix larger than WCHAR_MAX differences in wcs{,n}cmp (#795) The C standard states: > Unless explicitly stated otherwise, the functions described in this > subclause order two wide characters the same way as two integers of > the underlying integer type designated by wchar_t. > > [...] > > The wcscmp function returns an integer greater than, equal to, or > less than zero, accordingly as the wide string pointed to by s1 is > greater than, equal to, or less than the wide string pointed to by > s2. > > [...] > > The wcsncmp function returns an integer greater than, equal to, or > less than zero, accordingly as the possibly null-terminated array > pointed to by s1 is greater than, equal to, or less than the > possibly null-terminated array pointed to by s2. - C Standard, 7.31.4.4. Wide string comparison functions Cosmopolitan fails to obey this in cases where the difference between two wide characters is larger than WCHAR_MAX. This means that, for example, the following program: #include <stdio.h> #include <wchar.h> #include <limits.h> int main() { wchar_t str1[] = { WCHAR_MIN, L'\0' }; wchar_t str2[] = { WCHAR_MAX, L'\0' }; printf("%d\n", wcscmp(str1, str2)); printf("%d\n", wcsncmp(str1, str2, 2)); } will print `1` twice, instead of the negative numbers mandated by the standard (as WCHAR_MIN is less than WCHAR_MAX) This patch fixes this, along with the associated Github issue, https://github.com/jart/cosmopolitan/issues/783
2023-03-30 00:28:18 +00:00
EXPECT_LT(0, wcscmp(memcpy(B1, "\xff\xff\xff\x7f", 4),
memcpy(B2, "\x00\x00\x00\x80", 4)));
EXPECT_LT(wcscmp(memcpy(B1, "\x00\x00\x00\x80", 4),
Initial import
2020-06-15 14:18:57 +00:00
memcpy(B2, "\xff\xff\xff\x7f", 4)),
Fix larger than WCHAR_MAX differences in wcs{,n}cmp (#795) The C standard states: > Unless explicitly stated otherwise, the functions described in this > subclause order two wide characters the same way as two integers of > the underlying integer type designated by wchar_t. > > [...] > > The wcscmp function returns an integer greater than, equal to, or > less than zero, accordingly as the wide string pointed to by s1 is > greater than, equal to, or less than the wide string pointed to by > s2. > > [...] > > The wcsncmp function returns an integer greater than, equal to, or > less than zero, accordingly as the possibly null-terminated array > pointed to by s1 is greater than, equal to, or less than the > possibly null-terminated array pointed to by s2. - C Standard, 7.31.4.4. Wide string comparison functions Cosmopolitan fails to obey this in cases where the difference between two wide characters is larger than WCHAR_MAX. This means that, for example, the following program: #include <stdio.h> #include <wchar.h> #include <limits.h> int main() { wchar_t str1[] = { WCHAR_MIN, L'\0' }; wchar_t str2[] = { WCHAR_MAX, L'\0' }; printf("%d\n", wcscmp(str1, str2)); printf("%d\n", wcsncmp(str1, str2, 2)); } will print `1` twice, instead of the negative numbers mandated by the standard (as WCHAR_MIN is less than WCHAR_MAX) This patch fixes this, along with the associated Github issue, https://github.com/jart/cosmopolitan/issues/783
2023-03-30 00:28:18 +00:00
0);
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(B2);
free(B1);
Initial import
2020-06-15 14:18:57 +00:00
}
Make improvements
2020-12-01 11:43:40 +00:00
TEST(wcsncmp, testTwosComplementBane) {
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
wchar_t *B1 = malloc(4);
wchar_t *B2 = malloc(4);
Initial import
2020-06-15 14:18:57 +00:00
EXPECT_EQ(wcsncmp(memcpy(B1, "\x00\x00\x00\x80", 4),
memcpy(B2, "\x00\x00\x00\x80", 4), 1),
0);
Fix larger than WCHAR_MAX differences in wcs{,n}cmp (#795) The C standard states: > Unless explicitly stated otherwise, the functions described in this > subclause order two wide characters the same way as two integers of > the underlying integer type designated by wchar_t. > > [...] > > The wcscmp function returns an integer greater than, equal to, or > less than zero, accordingly as the wide string pointed to by s1 is > greater than, equal to, or less than the wide string pointed to by > s2. > > [...] > > The wcsncmp function returns an integer greater than, equal to, or > less than zero, accordingly as the possibly null-terminated array > pointed to by s1 is greater than, equal to, or less than the > possibly null-terminated array pointed to by s2. - C Standard, 7.31.4.4. Wide string comparison functions Cosmopolitan fails to obey this in cases where the difference between two wide characters is larger than WCHAR_MAX. This means that, for example, the following program: #include <stdio.h> #include <wchar.h> #include <limits.h> int main() { wchar_t str1[] = { WCHAR_MIN, L'\0' }; wchar_t str2[] = { WCHAR_MAX, L'\0' }; printf("%d\n", wcscmp(str1, str2)); printf("%d\n", wcsncmp(str1, str2, 2)); } will print `1` twice, instead of the negative numbers mandated by the standard (as WCHAR_MIN is less than WCHAR_MAX) This patch fixes this, along with the associated Github issue, https://github.com/jart/cosmopolitan/issues/783
2023-03-30 00:28:18 +00:00
EXPECT_GT(wcsncmp(memcpy(B1, "\xff\xff\xff\x7f", 4),
Initial import
2020-06-15 14:18:57 +00:00
memcpy(B2, "\x00\x00\x00\x80", 4), 1),
Fix larger than WCHAR_MAX differences in wcs{,n}cmp (#795) The C standard states: > Unless explicitly stated otherwise, the functions described in this > subclause order two wide characters the same way as two integers of > the underlying integer type designated by wchar_t. > > [...] > > The wcscmp function returns an integer greater than, equal to, or > less than zero, accordingly as the wide string pointed to by s1 is > greater than, equal to, or less than the wide string pointed to by > s2. > > [...] > > The wcsncmp function returns an integer greater than, equal to, or > less than zero, accordingly as the possibly null-terminated array > pointed to by s1 is greater than, equal to, or less than the > possibly null-terminated array pointed to by s2. - C Standard, 7.31.4.4. Wide string comparison functions Cosmopolitan fails to obey this in cases where the difference between two wide characters is larger than WCHAR_MAX. This means that, for example, the following program: #include <stdio.h> #include <wchar.h> #include <limits.h> int main() { wchar_t str1[] = { WCHAR_MIN, L'\0' }; wchar_t str2[] = { WCHAR_MAX, L'\0' }; printf("%d\n", wcscmp(str1, str2)); printf("%d\n", wcsncmp(str1, str2, 2)); } will print `1` twice, instead of the negative numbers mandated by the standard (as WCHAR_MIN is less than WCHAR_MAX) This patch fixes this, along with the associated Github issue, https://github.com/jart/cosmopolitan/issues/783
2023-03-30 00:28:18 +00:00
0);
EXPECT_LT(wcsncmp(memcpy(B1, "\x00\x00\x00\x80", 4),
Initial import
2020-06-15 14:18:57 +00:00
memcpy(B2, "\xff\xff\xff\x7f", 4), 1),
Fix larger than WCHAR_MAX differences in wcs{,n}cmp (#795) The C standard states: > Unless explicitly stated otherwise, the functions described in this > subclause order two wide characters the same way as two integers of > the underlying integer type designated by wchar_t. > > [...] > > The wcscmp function returns an integer greater than, equal to, or > less than zero, accordingly as the wide string pointed to by s1 is > greater than, equal to, or less than the wide string pointed to by > s2. > > [...] > > The wcsncmp function returns an integer greater than, equal to, or > less than zero, accordingly as the possibly null-terminated array > pointed to by s1 is greater than, equal to, or less than the > possibly null-terminated array pointed to by s2. - C Standard, 7.31.4.4. Wide string comparison functions Cosmopolitan fails to obey this in cases where the difference between two wide characters is larger than WCHAR_MAX. This means that, for example, the following program: #include <stdio.h> #include <wchar.h> #include <limits.h> int main() { wchar_t str1[] = { WCHAR_MIN, L'\0' }; wchar_t str2[] = { WCHAR_MAX, L'\0' }; printf("%d\n", wcscmp(str1, str2)); printf("%d\n", wcsncmp(str1, str2, 2)); } will print `1` twice, instead of the negative numbers mandated by the standard (as WCHAR_MIN is less than WCHAR_MAX) This patch fixes this, along with the associated Github issue, https://github.com/jart/cosmopolitan/issues/783
2023-03-30 00:28:18 +00:00
0);
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
free(B2);
free(B1);
Initial import
2020-06-15 14:18:57 +00:00
}
/*───────────────────────────────────────────────────────────────────────────│─╗
│ test/libc/str/strcmp_test.c § benchmarks ─╬─│┼
╚────────────────────────────────────────────────────────────────────────────│*/
Remove `testonly` keyword
2022-09-05 15:41:43 +00:00
dontinline int strcmp_pure(const char *a, const char *b) {
Initial import
2020-06-15 14:18:57 +00:00
for (; *a == *b; a++, b++) {
if (!*a) break;
}
return (*a & 0xff) - (*b & 0xff);
}
Remove `testonly` keyword
2022-09-05 15:41:43 +00:00
dontinline int strcasecmp_pure(const char *a, const char *b) {
Initial import
2020-06-15 14:18:57 +00:00
for (; *a && *b; a++, b++) {
if (!(*a == *b || tolower(*a & 0xff) == tolower(*b & 0xff))) {
break;
}
}
return tolower(*a & 0xff) - tolower(*b & 0xff);
}
Fix some build errors
2023-06-05 22:50:15 +00:00
char *randomize_buf2str(size_t size, char *data) {
assert(data);
Rename rand64() to _rand64()
2022-10-10 11:12:06 +00:00
rngset(data, size, _rand64, -1);
Initial import
2020-06-15 14:18:57 +00:00
data[size - 1] = '\0';
return data;
}
Fix some build errors
2023-06-05 22:50:15 +00:00
char *longstringislong(size_t size, char *data) {
Initial import
2020-06-15 14:18:57 +00:00
unsigned i;
Fix some build errors
2023-06-05 22:50:15 +00:00
assert(data);
Initial import
2020-06-15 14:18:57 +00:00
randomize_buf2str(size, data);
for (i = 0; i < size; ++i) {
data[i] |= 1u << (i & 5);
}
data[size - 1] = '\0';
return data;
}
void randomize_buf2str_dupe(size_t size, char data[size], char dupe[size]) {
randomize_buf2str(size, data);
memcpy(dupe, data, size);
}
void longstringislong_dupe(size_t size, char data[size], char dupe[size]) {
longstringislong(size, data);
memcpy(dupe, data, size);
}
BENCH(bench_00_strcmp, bench) {
size_t size;
char *dupe, *data;
Add automatic TMPDIR setup/teardown to GNU Make We now guarantee TMPDIR will be defined on a per build rule basis. It'll be an absolute path. It'll be secure and unique. It'll be rm -rf'd after the last shell script line in your build rule is executed. If $TMPDIR is already defined, then it'll be created as a subdirectory of your $TMPDIR and then replace the variable with the new definition. The Landlock Make repository will be updated with examples shortly after this change which shall be known as Landlock Make 1.1.1. See #530
2022-08-14 08:21:26 +00:00
size = 14139;
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
data = gc(malloc(size));
dupe = gc(malloc(size));
Make improvements
2020-12-01 11:43:40 +00:00
Initial import
2020-06-15 14:18:57 +00:00
EZBENCH2("strcmp [identity]", longstringislong(size, data),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp(__veil("r", data), data)));
Make improvements
2020-12-01 11:43:40 +00:00
EZBENCH2("strcmp [2 diff]", donothing,
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp(__veil("r", "hi"), __veil("r", "there"))));
Perform some code cleanup
2023-05-15 23:32:10 +00:00
EZBENCH2("scmppure [2 diff]", donothing,
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp_pure(__veil("r", "hi"), __veil("r", "there"))));
Make improvements
2020-12-01 11:43:40 +00:00
EZBENCH2("strcmp [2 dupe]", randomize_buf2str_dupe(2, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp(__veil("r", data), __veil("r", dupe))));
Perform some code cleanup
2023-05-15 23:32:10 +00:00
EZBENCH2("scmp_pure [2 dupe]", randomize_buf2str_dupe(2, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp_pure(__veil("r", data), __veil("r", dupe))));
Make improvements
2020-12-01 11:43:40 +00:00
EZBENCH2("strcmp [4 dupe]", randomize_buf2str_dupe(4, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp(__veil("r", data), __veil("r", dupe))));
Perform some code cleanup
2023-05-15 23:32:10 +00:00
EZBENCH2("scmp_pure [4 dupe]", randomize_buf2str_dupe(4, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp_pure(__veil("r", data), __veil("r", dupe))));
Make improvements
2020-12-01 11:43:40 +00:00
EZBENCH2("strcmp [8 dupe]", randomize_buf2str_dupe(8, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp(__veil("r", data), __veil("r", dupe))));
Perform some code cleanup
2023-05-15 23:32:10 +00:00
EZBENCH2("scmp_pure [8 dupe]", randomize_buf2str_dupe(8, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp_pure(__veil("r", data), __veil("r", dupe))));
Make improvements
2020-12-01 11:43:40 +00:00
Get more Python tests passing (#141)
2021-08-16 22:26:31 +00:00
EZBENCH2("strcmp [sdupe]", randomize_buf2str_dupe(size, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp(__veil("r", data), __veil("r", dupe))));
Perform some code cleanup
2023-05-15 23:32:10 +00:00
EZBENCH2("scmp_pure [sdupe]", randomize_buf2str_dupe(size, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp_pure(__veil("r", data), __veil("r", dupe))));
Make improvements
2020-12-01 11:43:40 +00:00
Get more Python tests passing (#141)
2021-08-16 22:26:31 +00:00
EZBENCH2("strcmp [ldupe]", longstringislong_dupe(size, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp(__veil("r", data), __veil("r", dupe))));
Perform some code cleanup
2023-05-15 23:32:10 +00:00
EZBENCH2("scmp_pure [ldupe]", longstringislong_dupe(size, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcmp_pure(__veil("r", data), __veil("r", dupe))));
Initial import
2020-06-15 14:18:57 +00:00
}
BENCH(bench_01_strcasecmp, bench) {
size_t size;
char *dupe, *data;
Add automatic TMPDIR setup/teardown to GNU Make We now guarantee TMPDIR will be defined on a per build rule basis. It'll be an absolute path. It'll be secure and unique. It'll be rm -rf'd after the last shell script line in your build rule is executed. If $TMPDIR is already defined, then it'll be created as a subdirectory of your $TMPDIR and then replace the variable with the new definition. The Landlock Make repository will be updated with examples shortly after this change which shall be known as Landlock Make 1.1.1. See #530
2022-08-14 08:21:26 +00:00
size = 141393;
Make C memory safe like Rust This change enables Address Sanitizer systemically w/ `make MODE=dbg`. Our version of Rust's `unsafe` keyword is named `noasan` which is used for two functions that do aligned memory chunking, like `strcpy.c` and we need to fix the tiny DEFLATE code, but that's it everything else is fabulous you can have all the fischer price security blankets you need Best of all is we're now able to use the ASAN data in Blinkenlights to colorize the memory dumps. See the screenshot below of a test program: https://justine.lol/blinkenlights/asan.png Which is operating on float arrays stored on the stack, with red areas indicating poisoned memory, and the green areas indicate valid memory.
2021-02-01 11:33:13 +00:00
data = gc(malloc(size));
dupe = gc(malloc(size));
Make improvements
2020-12-01 11:43:40 +00:00
Perform some code cleanup
2023-05-15 23:32:10 +00:00
EZBENCH2("strcasecmp [iden]", longstringislong(size, data),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcasecmp(__veil("r", data), data)));
Make improvements
2020-12-01 11:43:40 +00:00
Get more Python tests passing (#141)
2021-08-16 22:26:31 +00:00
EZBENCH2("strcasecmp [sdupe]", randomize_buf2str_dupe(size, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcasecmp(__veil("r", data), __veil("r", dupe))));
EZBENCH2(
"sccmp_pure [sdupe]", randomize_buf2str_dupe(size, data, dupe),
__expropriate(strcasecmp_pure(__veil("r", data), __veil("r", dupe))));
Make improvements
2020-12-01 11:43:40 +00:00
Get more Python tests passing (#141)
2021-08-16 22:26:31 +00:00
EZBENCH2("strcasecmp [ldupe]", longstringislong_dupe(size, data, dupe),
Mint APE Loader v1.5 This change ports APE Loader to Linux AARCH64, so that Raspberry Pi users can run programs like redbean, without the executable needing to modify itself. Progress has also slipped into this change on the issue of making progress better conforming to user expectations and industry standards regarding which symbols we're allowed to declare
2023-07-26 20:54:49 +00:00
__expropriate(strcasecmp(__veil("r", data), __veil("r", dupe))));
EZBENCH2(
"sccmp_pure [ldupe]", longstringislong_dupe(size, data, dupe),
__expropriate(strcasecmp_pure(__veil("r", data), __veil("r", dupe))));
Initial import
2020-06-15 14:18:57 +00:00
}
Add SNI support to redbean and improve SSL perf This change makes SSL virtual hosting possible. You can now load multiple certificates for multiple domains and redbean will just figure out which one to use, even if you only have 1 ip address. You can also use a jumbo certificate that lists all your domains in the the subject alternative names. This change also makes performance improvements to MbedTLS. Here are some benchmarks vs. cc1920749eb81f346badaf55fbf79620cb718a55 BEFORE AFTER (microsecs) suite_ssl.com 2512881 191738 13.11x faster suite_pkparse.com 36291 3295 11.01x faster suite_x509parse.com 854669 120293 7.10x faster suite_pkwrite.com 6549 1265 5.18x faster suite_ecdsa.com 53347 18778 2.84x faster suite_pk.com 49051 18717 2.62x faster suite_ecdh.com 19535 9502 2.06x faster suite_shax.com 15848 7965 1.99x faster suite_rsa.com 353257 184828 1.91x faster suite_x509write.com 162646 85733 1.90x faster suite_ecp.com 20503 11050 1.86x faster suite_hmac_drbg.no_reseed.com 19528 11417 1.71x faster suite_hmac_drbg.nopr.com 12460 8010 1.56x faster suite_mpi.com 687124 442661 1.55x faster suite_hmac_drbg.pr.com 11890 7752 1.53x faster There aren't any special tricks to the performance imporvements. It's mostly due to code cleanup, assembly and intel instructions like mulx, adox, and adcx.
2021-07-19 21:55:20 +00:00
BENCH(memcmp, bench) {
volatile char *copy = gc(strdup(kHyperion));
EZBENCH2("memcmp big", donothing,
Fix warnings This change fixes Cosmopolitan so it has fewer opinions about compiler warnings. The whole repository had to be cleaned up to be buildable in -Werror -Wall mode. This lets us benefit from things like strict const checking. Some actual bugs might have been caught too.
2023-09-02 03:49:13 +00:00
__expropriate(memcmp(kHyperion, (void *)copy, kHyperionSize)));
Correct more typos (#500)
2022-07-22 03:53:30 +00:00
copy = gc(strdup("tough little ship"));
Reduce build graph by another eight percent
2022-08-13 20:11:56 +00:00
EZBENCH2("memcmp 18", donothing,
Fix warnings This change fixes Cosmopolitan so it has fewer opinions about compiler warnings. The whole repository had to be cleaned up to be buildable in -Werror -Wall mode. This lets us benefit from things like strict const checking. Some actual bugs might have been caught too.
2023-09-02 03:49:13 +00:00
__expropriate(memcmp("tough little ship", (void *)copy, 18)));
Add SNI support to redbean and improve SSL perf This change makes SSL virtual hosting possible. You can now load multiple certificates for multiple domains and redbean will just figure out which one to use, even if you only have 1 ip address. You can also use a jumbo certificate that lists all your domains in the the subject alternative names. This change also makes performance improvements to MbedTLS. Here are some benchmarks vs. cc1920749eb81f346badaf55fbf79620cb718a55 BEFORE AFTER (microsecs) suite_ssl.com 2512881 191738 13.11x faster suite_pkparse.com 36291 3295 11.01x faster suite_x509parse.com 854669 120293 7.10x faster suite_pkwrite.com 6549 1265 5.18x faster suite_ecdsa.com 53347 18778 2.84x faster suite_pk.com 49051 18717 2.62x faster suite_ecdh.com 19535 9502 2.06x faster suite_shax.com 15848 7965 1.99x faster suite_rsa.com 353257 184828 1.91x faster suite_x509write.com 162646 85733 1.90x faster suite_ecp.com 20503 11050 1.86x faster suite_hmac_drbg.no_reseed.com 19528 11417 1.71x faster suite_hmac_drbg.nopr.com 12460 8010 1.56x faster suite_mpi.com 687124 442661 1.55x faster suite_hmac_drbg.pr.com 11890 7752 1.53x faster There aren't any special tricks to the performance imporvements. It's mostly due to code cleanup, assembly and intel instructions like mulx, adox, and adcx.
2021-07-19 21:55:20 +00:00
}
Perform some code cleanup
2023-05-15 23:32:10 +00:00
/* jart
* strcmp [identity] l: 3c 1ns m: 30c 10ns
* strcmp [2 diff] l: 4c 1ns m: 30c 10ns
* scmppure [2 diff] l: 3c 1ns m: 31c 10ns
* strcmp [2 dupe] l: 8c 3ns m: 39c 13ns
* scmp_pure [2 dupe] l: 6c 2ns m: 34c 11ns
* strcmp [4 dupe] l: 9c 3ns m: 40c 13ns
* scmp_pure [4 dupe] l: 9c 3ns m: 38c 12ns
* strcmp [8 dupe] l: 10c 3ns m: 40c 13ns
* scmp_pure [8 dupe] l: 11c 4ns m: 41c 13ns
* strcmp [sdupe] l: 87c 28ns m: 121c 39ns
* scmp_pure [sdupe] l: 188c 61ns m: 294c 95ns
* strcmp [ldupe] l: 3,458c 1,117ns m: 3,486c 1,126ns
* scmp_pure [ldupe] l: 11,441c 3,695ns m: 11,520c 3,721ns
* strcasecmp [iden] l: 3c 1ns m: 30c 10ns
* strcasecmp [sdupe] l: 105c 34ns m: 156c 50ns
* sccmp_pure [sdupe] l: 644c 208ns m: 963c 311ns
* strcasecmp [ldupe] l: 36,527c 11,798ns m: 36,582c 11,816ns
* sccmp_pure [ldupe] l: 365,880c 118,177ns m: 365,721c 118,125ns
* memcmp big l: 2,050c 662ns m: 2,093c 676ns
* memcmp 18 l: 6c 2ns m: 35c 11ns
*/
/* jart+intel
* strcmp [identity] l: 1c 0ns m: 28c 9ns
* strcmp [2 diff] l: 2c 1ns m: 29c 9ns
* scmppure [2 diff] l: 2c 1ns m: 29c 9ns
* strcmp [2 dupe] l: 8c 3ns m: 40c 13ns
* scmp_pure [2 dupe] l: 5c 2ns m: 32c 10ns
* strcmp [4 dupe] l: 9c 3ns m: 41c 13ns
* scmp_pure [4 dupe] l: 7c 2ns m: 34c 11ns
* strcmp [8 dupe] l: 10c 3ns m: 40c 13ns
* scmp_pure [8 dupe] l: 10c 3ns m: 39c 13ns
* strcmp [sdupe] l: 57c 18ns m: 87c 28ns
* scmp_pure [sdupe] l: 191c 62ns m: 224c 72ns
* strcmp [ldupe] l: 1,667c 538ns m: 1,708c 552ns
* scmp_pure [ldupe] l: 10,988c 3,549ns m: 11,055c 3,571ns
* strcasecmp [iden] l: 2c 1ns m: 31c 10ns
* strcasecmp [sdupe] l: 121c 39ns m: 160c 52ns
* sccmp_pure [sdupe] l: 684c 221ns m: 702c 227ns
* strcasecmp [ldupe] l: 34,325c 11,087ns m: 35,954c 11,613ns
* sccmp_pure [ldupe] l: 361,449c 116,746ns m: 366,022c 118,223ns
* memcmp big l: 2,040c 659ns m: 2,083c 673ns
* memcmp 18 l: 5c 2ns m: 35c 11ns
*/
Reference in a new issue Copy permalink