busybox-w32/scripts, branch masterA mirror of https://github.com/rmyorston/busybox-w32.git
https://git.lua4.win/busybox-w32/atom?h=master2026-04-18T09:27:47+00:00Win32: trivial refinements (#577)2026-04-18T09:27:47+00:00avih2164962+avih@users.noreply.github.com2026-04-18T09:27:47+00:00urn:sha1:0495eb5c75d62fc574a8c5477f9196f0917f723d
* win32: fnmatch2 minor refinements (no-op)
With beacket, add '\' to the switch/case, and at fnmatch, pre-calc
the two most used flags (CASEFOLD and PATHNAME) once on init.
Negligible perf impact, and happens to save few bytes in x64.
* win32: actype: trivial refinments (no-op)
NULL was used in actype.h but requires stddef.h - which we don't
include, so use (int*)0 which also expresses it more clearly.
Remove "#include <string.h>" at actype.c, which we don't need anymore.
* scripts/patbench.sh: make executable (chmod +x)scripts/patbench.sh: add patterns, verify results2026-04-16T09:42:13+00:00Avi Halachmi (:avih)avihpit@yahoo.com2026-04-15T15:51:27+00:00urn:sha1:797628de79fe79e95da3e783b10a642c6e03c234
Few minor improvements:
- Add few valid but edge-case patterns (start with ']' etc).
- Add the expected result for default lines, and report ok/err.
- Disable measurements with -mn (to be measured externally).
Also, the default lines were modified slightly so that the new
edge cases results are largely distinct, so it can be used as
correctness test with -d (and exit code is 1 on errors).
BENCHMARK RESULTS
Below are results of running this script using busybox-w32 on win10,
and using busybox on Alpine linux 3.23.
Observations of busybox-w32 results, which uses old glibc fnmatch:
[10] 129 ms OK (M:12345) '*[hello world, this is a test]*'
[42] 22491 ms OK (M:12-45) '*[0-9A-Za-z]*[0-9A-Za-z]*[0-9A-Za-z]'
The current busybox-w32 implementation is exponential, and while
tests of simple patterns complete in 100-200ms, patterns with 3 '*'
can take 30s or more to complete, way longer with 4 '*', etc.
[12] 192 ms E:1-3-- (M:-----) '*[*'
[13] 201 ms E:1---- (M:-----) '*[]*'
[14] 203 ms E:--3-- (M:-----) '*[!]*'
A bracket which doesn't start a bracket expression is regular char,
and all these are missing the closing bracket (by POSIX rules), and
should be matched literally, but the current fnmatch fails with that.
[35] 1305 ms OK (M:12-45) '*[0-9A-Za-z]*[0-9A-Za-z]'
[36] 1928 ms OK (M:12-45) '*[[:alnum:]]*[[:alnum:]]'
[38] 1446 ms OK (M:12-45) '*[a-bc-de-z]*[a-bc-de-z]'
[39] 3192 ms OK (M:12-45) '*[[:lower:]]*[[:lower:]]'
[40] 1586 ms OK (M:---4-) '*[0-9A-Fa-ef]*[0-9A-Fa-ef]'
[41] 4601 ms OK (M:---4-) '*[[:xdigit:]]*[[:xdigit:]]'
charclass [:alnum:] is x1.4 slower than an equivalent non-class range,
but [:lower:] is already x2 slower than the equivalent, and [:xdigit:]
is x3 slower than the equivalent. This is because the name is searched
in a list which starts with "alnum" and ends in "xdigit", and the
search duration (and the following switch in the same order) clearly
affects how quickly a pattern can be tested.
There are other issues in busybox-w32 fnmatch, like incorrect
case-insensitivity and more, but which don't manifest in case...esac.
Benchmark results with busybox-w32 on Windows 10 (compiled with gcc):
[ 1] 107 ms OK (M:-----) ''
[ 2] 108 ms OK (M:12345) '*'
[ 3] 104 ms OK (M:-----) '@'
[ 4] 112 ms OK (M:1----) 'Lorem*'
[ 5] 120 ms OK (M:-2---) '*quis'
[ 6] 127 ms OK (M:--3--) '*nisi*'
[ 7] 244 ms OK (M:--3--) '*[fobar]xyz*'
[ 8] 544 ms OK (M:--3--) '*[foobarfoobarfoobarfoobar]xyz*'
[ 9] 588 ms OK (M:-2---) '*[!foobarfoobarfoobarfoobar]xyz*'
[10] 129 ms OK (M:12345) '*[hello world, this is a test]*'
[11] 136 ms OK (M:12345) '*[!hello world, this is a test]*'
[12] 192 ms E:1-3-- (M:-----) '*[*'
[13] 201 ms E:1---- (M:-----) '*[]*'
[14] 203 ms E:--3-- (M:-----) '*[!]*'
[15] 182 ms OK (M:-23-5) '*[!]].*'
[16] 454 ms OK (M:1----) '*[![:print:]]*'
[17] 124 ms OK (M:-23--) '*z*'
[18] 123 ms OK (M:---4-) '*-*'
[19] 160 ms OK (M:---4-) '*[-]*'
[20] 175 ms OK (M:-234-) '*[-z]*'
[21] 177 ms OK (M:-234-) '*[z-]*'
[22] 197 ms OK (M:-234-) '*[-z-]*'
[23] 126 ms OK (M:1-3--) '*]*'
[24] 168 ms OK (M:1-3--) '*[]]*'
[25] 173 ms OK (M:1-34-) '*[]-]*'
[26] 168 ms OK (M:123--) '*[]-_]*'
[27] 170 ms OK (M:1234-) '*[]-_-]*'
[28] 365 ms OK (M:12-45) '*[0-9A-Za-z][0-9A-Za-z]'
[29] 513 ms OK (M:12-45) '*[[:alnum:]][[:alnum:]]'
[30] 263 ms OK (M:12-45) '*[a-z][a-z]'
[31] 376 ms OK (M:12-45) '*[a-bc-de-z][a-bc-de-z]'
[32] 815 ms OK (M:12-45) '*[[:lower:]][[:lower:]]'
[33] 311 ms OK (M:-----) '*[0-9A-Fa-ef][0-9A-Fa-ef]'
[34] 756 ms OK (M:-----) '*[[:xdigit:]][[:xdigit:]]'
[35] 1305 ms OK (M:12-45) '*[0-9A-Za-z]*[0-9A-Za-z]'
[36] 1928 ms OK (M:12-45) '*[[:alnum:]]*[[:alnum:]]'
[37] 919 ms OK (M:12-45) '*[a-z]*[a-z]'
[38] 1446 ms OK (M:12-45) '*[a-bc-de-z]*[a-bc-de-z]'
[39] 3192 ms OK (M:12-45) '*[[:lower:]]*[[:lower:]]'
[40] 1586 ms OK (M:---4-) '*[0-9A-Fa-ef]*[0-9A-Fa-ef]'
[41] 4601 ms OK (M:---4-) '*[[:xdigit:]]*[[:xdigit:]]'
[42] 22491 ms OK (M:12-45) '*[0-9A-Za-z]*[0-9A-Za-z]*[0-9A-Za-z]'
(aborted manually - exponential implementation becomes very slow)
results with busybox on Alpine linux 3.23 (musl-libc has linear-time
fnmatch, using wide-char wctype/iswctype - slightly slower than char):
[ 1] 135 ms OK (M:-----) ''
[ 2] 141 ms OK (M:12345) '*'
[ 3] 139 ms OK (M:-----) '@'
[ 4] 141 ms OK (M:1----) 'Lorem*'
[ 5] 148 ms OK (M:-2---) '*quis'
[ 6] 216 ms OK (M:--3--) '*nisi*'
[ 7] 410 ms OK (M:--3--) '*[fobar]xyz*'
[ 8] 787 ms OK (M:--3--) '*[foobarfoobarfoobarfoobar]xyz*'
[ 9] 844 ms OK (M:-2---) '*[!foobarfoobarfoobarfoobar]xyz*'
[10] 215 ms OK (M:12345) '*[hello world, this is a test]*'
[11] 237 ms OK (M:12345) '*[!hello world, this is a test]*'
[12] 206 ms OK (M:1-3--) '*[*'
[13] 242 ms OK (M:1----) '*[]*'
[14] 220 ms OK (M:--3--) '*[!]*'
[15] 306 ms OK (M:-23-5) '*[!]].*'
[16] 591 ms OK (M:1----) '*[![:print:]]*'
[17] 216 ms OK (M:-23--) '*z*'
[18] 198 ms OK (M:---4-) '*-*'
[19] 286 ms OK (M:---4-) '*[-]*'
[20] 285 ms OK (M:-234-) '*[-z]*'
[21] 291 ms OK (M:-234-) '*[z-]*'
[22] 299 ms OK (M:-234-) '*[-z-]*'
[23] 228 ms OK (M:1-3--) '*]*'
[24] 286 ms OK (M:1-3--) '*[]]*'
[25] 267 ms OK (M:1-34-) '*[]-]*'
[26] 349 ms OK (M:123--) '*[]-_]*'
[27] 292 ms OK (M:1234-) '*[]-_-]*'
[28] 198 ms OK (M:12-45) '*[0-9A-Za-z][0-9A-Za-z]'
[29] 201 ms OK (M:12-45) '*[[:alnum:]][[:alnum:]]'
[30] 179 ms OK (M:12-45) '*[a-z][a-z]'
[31] 217 ms OK (M:12-45) '*[a-bc-de-z][a-bc-de-z]'
[32] 205 ms OK (M:12-45) '*[[:lower:]][[:lower:]]'
[33] 202 ms OK (M:-----) '*[0-9A-Fa-ef][0-9A-Fa-ef]'
[34] 203 ms OK (M:-----) '*[[:xdigit:]][[:xdigit:]]'
[35] 215 ms OK (M:12-45) '*[0-9A-Za-z]*[0-9A-Za-z]'
[36] 213 ms OK (M:12-45) '*[[:alnum:]]*[[:alnum:]]'
[37] 191 ms OK (M:12-45) '*[a-z]*[a-z]'
[38] 219 ms OK (M:12-45) '*[a-bc-de-z]*[a-bc-de-z]'
[39] 224 ms OK (M:12-45) '*[[:lower:]]*[[:lower:]]'
[40] 217 ms OK (M:---4-) '*[0-9A-Fa-ef]*[0-9A-Fa-ef]'
[41] 199 ms OK (M:---4-) '*[[:xdigit:]]*[[:xdigit:]]'
[42] 227 ms OK (M:12-45) '*[0-9A-Za-z]*[0-9A-Za-z]*[0-9A-Za-z]'
[43] 239 ms OK (M:12-45) '*[[:alnum:]]*[[:alnum:]]*[[:alnum:]]'
[44] 204 ms OK (M:12-45) '*[a-z]*[a-z]*[a-z]'
[45] 258 ms OK (M:12-45) '*[a-bc-de-z]*[a-bc-de-z]*[a-bc-de-z]'
[46] 235 ms OK (M:12-45) '*[[:lower:]]*[[:lower:]]*[[:lower:]]'
[47] 247 ms OK (M:---4-) '*[0-9A-Fa-ef]*[0-9A-Fa-ef]*[0-9A-Fa-ef]'
[48] 247 ms OK (M:---4-) '*[[:xdigit:]]*[[:xdigit:]]*[[:xdigit:]]'
[49] 286 ms OK (M:12-45) '*[0-9A-Za-z]*[0-9A-Za-z]*[0-9A-Za-z]*[0-9A-Za-z]'
[50] 262 ms OK (M:12-45) '*[[:alnum:]]*[[:alnum:]]*[[:alnum:]]*[[:alnum:]]'
[51] 219 ms OK (M:12-45) '*[a-z]*[a-z]*[a-z]*[a-z]'
[52] 273 ms OK (M:12-45) '*[a-bc-de-z]*[a-bc-de-z]*[a-bc-de-z]*[a-bc-de-z]'
[53] 274 ms OK (M:12-45) '*[[:lower:]]*[[:lower:]]*[[:lower:]]*[[:lower:]]'
[54] 281 ms OK (M:---4-) '*[0-9A-Fa-ef]*[0-9A-Fa-ef]*[0-9A-Fa-ef]*[0-9A-Fa-ef]'
[55] 296 ms OK (M:---4-) '*[[:xdigit:]]*[[:xdigit:]]*[[:xdigit:]]*[[:xdigit:]]'
scripts/patbench.sh: new shell patterns benchmark2026-04-16T09:42:13+00:00Avi Halachmi (:avih)avihpit@yahoo.com2026-04-12T04:45:25+00:00urn:sha1:0954c248c879e3859c5e9da573979be643d9f2ab
This script measures the performance of shell pattern matching,
with increasing pattern complexity (more '*' at the pattern).
Merge branch 'busybox' into merge2026-03-11T10:14:49+00:00Ron Yorstonrmy@pobox.com2026-03-11T10:14:49+00:00urn:sha1:ede205bd07573813337b5706acb0cae3b127a36f*: placate warnings where strchr/strstr returns constant pointer2026-02-15T14:15:30+00:00Denys Vlasenkovda.linux@googlemail.com2026-02-15T13:41:20+00:00urn:sha1:b668e52c906b664b353d5a99cfa3ff36f73b341d
Newer glibc is now smarter and can propagate const-ness from those!
function old new delta
readtoken1 3111 3108 -3
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
win32: wcwidth_alt.c: new generator script, re-generate code2025-12-29T13:57:05+00:00Avi Halachmi (:avih)avihpit@yahoo.com2025-12-25T22:02:45+00:00urn:sha1:6f847c296bf94401c0c147bf3c5aa1e8531b9dbf
Previously, scripts/mkwcwidth generated a wcwidth implementation based
on input from the python wcwidth: https://github.com/jquast/wcwidth .
However, that project is not updated frequently, and the less
dependencies - the better.
This new script instead prints an implementation based directly on the
original Unicode files - which it can also download.
It generates the same code as before (same data structures, etc).
It was then used to re-generate libbb/wcwidth_alt.c, as follows:
./scripts/mkwcwidth DL=15.1.0 FAST_FUNC > libbb/wcwidth_alt.c
This is the same Unicode version as the previous file, just to make
the codepoint differences visible easily.
These differences are very small, and the script itself explains where
and why it differs from https://github.com/jquast/wcwidth .
Next commits will update the tables to the latest unicode version.
Use Windows library for cryptographic checksums2025-06-09T12:23:49+00:00Ron Yorstonrmy@pobox.com2025-06-09T12:23:49+00:00urn:sha1:596c443112d09506c3bf13ac98046a84b912e56c
Add a new feature to libbb, FEATURE_USE_CNG_API, which enables the
use of the Cryptography API: Next Generation library to calculate
checksums. It is disabled by default except in the mingw64u default
config, as the API requires Windows 10+ to function. Usage of this
API provides a size benefit and delegates hardware optimizations to
the operating system cryptography library.
Based on GitHub PR #498 by rfl890.
Saves 4064 bytes in the mingw64u case.
Merge branch 'busybox' into merge2025-05-19T07:34:32+00:00Ron Yorstonrmy@pobox.com2025-05-19T07:34:32+00:00urn:sha1:64bf69893bd99c305d13a956389f216e7d15c682archival: disallow path traversals (CVE-2023-39810)2025-04-16T01:03:17+00:00Denys Vlasenkovda.linux@googlemail.com2024-10-02T08:12:05+00:00urn:sha1:9a8796436b9b0641e13480811902ea2ac57881d3
Create new configure option for archival/libarchive based extractions to
disallow path traversals.
As this is a paranoid option and might introduce backward
incompatibility, default it to no.
Fixes: CVE-2023-39810
Based on the patch by Peter Kaestle <peter.kaestle@nokia.com>
function old new delta
data_extract_all 921 945 +24
strip_unsafe_prefix 101 102 +1
------------------------------------------------------------------------------
(add/remove: 0/0 grow/shrink: 2/0 up/down: 25/0) Total: 25 bytes
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Update PDCurses2025-04-10T18:41:43+00:00Ron Yorstonrmy@pobox.com2025-04-10T18:41:43+00:00urn:sha1:486c1fcc341472fd8d64b820c3a7934cf3f42132