aboutsummaryrefslogtreecommitdiff
Commit message (Expand)AuthorAgeFilesLines
* Rewrite code, but keeping the exact same semantic and behaviour•••- assert that the generated `siphash.o` code is identical to the original code, available at `tests/assert-identical.sh`; - remove unneeded `#define`s; - rewrite code with the correct indentation, spacing and formatting; - use C99 constructs over C89 (for loop variable declarations inside the parentheses); - fix the public API. HEADmainEuAndreh2025-06-0425-3153/+858
* Add complete "Makefile" for standard packaging•••Also include a ".gitignore" with the derived assets. --- On the "Makefile", the canonical virtual targets implemented are: - "all": the default target name when no explicit name is given, or when one wants to build more than one target at once, such as: `make all check`. All it builds is the "libsiphash.a" library; - "check": NOOP, as there are no tests implemented; - "clean": removes the "*.o", "*.bin" and "libsiphash.a" files, which represent 100% of the generated assets; - "install": uses `$(DESTDIR)`, `$(LIBDIR)`, `$(INCLUDEDIR)` and `$(SRCDIR)` to properly place the "libsiphash.a" library, the "siphash.h" header and the "siphash.c" source in the correct place. `$(LIBDIR)`, `$(INCLUDEDIR)` and `$(SRCDIR)` are defined based on `$(PREFIX)`; - "uninstall": a perfect mirror of "install", which removes 100% of the installed artifacts. EuAndreh2025-03-202-0/+47
* Remove support filesEuAndreh2025-03-205-420/+0
* xJP2024-05-171-1/+1
* LLVM-compatible licenseJP2024-05-172-1/+225
* ipJP Aumasson2023-02-251-0/+4
* commentsJP Aumasson2022-10-223-3/+19
* duallicenseJP Aumasson2022-10-223-0/+9
* Merge pull request #46 from devendranaga/master•••SipHash works on ESP8266 running FreeRTOS. So Update the README on OS…Jean-Philippe Aumasson2022-10-221-1/+1
|\
| * SipHash works on ESP8266 running FreeRTOS. So Update the README on OS supportDev2022-08-171-1/+1
|/
* debug symbol fixJP Aumasson2022-07-281-1/+1
* Merge pull request #39 from snej/rename-debug-symbol•••Rename `DEBUG` to `DEBUG_SIPHASH` – fixing the makefile upstreamJean-Philippe Aumasson2022-07-282-2/+5
|\
| * Merge branch 'master' into rename-debug-symbolJean-Philippe Aumasson2022-07-283-3/+8
| |\ | |/ |/|
* | Merge pull request #41 from noocsharp/master•••fix typo: clarify -> clarityJean-Philippe Aumasson2022-01-211-1/+1
|\ \
| * | fix typo: clarify -> clarityNihal Jere2022-01-201-1/+1
|/ /
* | Merge pull request #40 from noocsharp/master•••only include stdio when usedJean-Philippe Aumasson2022-01-142-2/+4
|\ \
| * | only include stdio when usedNihal Jere2022-01-122-2/+4
|/ /
| * Rename `DEBUG` to `DEBUG_SIPHASH`•••`DEBUG` is too generic a name to use for a flag that causes SipHash to log to the console on every hash operation; changed it to `DEBUG_SIPHASH`. An example: Xcode's project templates have a `-DDEBUG=1` Clang flag in the 'Debug' config, so nearly every iOS and macOS app defines DEBUG in dev builds! When I (@snej) added SipHash to my Mac project, I started getting huge amounts of console logs until I found and renamed its `#ifdef DEBUG`. Jens Alfke2021-11-112-2/+2
|/
* httpsJP Aumasson2021-10-311-1/+1
* halfsh header, cleanupJP Aumasson2021-10-311-2/+1
* Merge pull request #37 from pjanx/master•••-Wimplicit-fallthrough=4 complianceJean-Philippe Aumasson2021-10-312-0/+8
|\
| * -Wimplicit-fallthrough=4 compliancePřemysl Eric Janouch2021-10-132-0/+8
|/
* Merge pull request #36 from themaxdavitt/themaxdavitt-patch-1•••Fix link to hash-flooding DOS attack slidesJean-Philippe Aumasson2021-06-071-1/+1
|\
| * Fix link to hash-flooding DOS attack slidesMax Davitt2021-06-061-1/+1
|/
* Merge pull request #35 from veorq/uint8t2uchar•••address potential strict aliasing violationJean-Philippe Aumasson2021-03-126-31/+36
|\
| * constsJP Aumasson2021-03-122-6/+6
| * typechangeJP Aumasson2021-02-206-31/+36
|/
* Merge pull request #34 from veorq/moretest•••more tests and security checksJean-Philippe Aumasson2021-02-197-350/+2348
|\
| * testsJP Aumasson2021-02-197-350/+2348
|/
* halfsiphash header + badgeJP Aumasson2021-02-194-3/+29
* Merge pull request #32 from veorq/header•••headerJean-Philippe Aumasson2021-01-192-3/+27
|\
| * fixesJP Aumasson2021-01-192-59/+62
| * headerJP Aumasson2021-01-192-61/+82
|/
* Merge pull request #30 from veorq/notahash•••NotahashJean-Philippe Aumasson2021-01-114-42/+77
|\
| * ignore binsJP Aumasson2021-01-111-0/+5
| * formattingJean-Philippe Aumasson2021-01-111-3/+3
| * readmeJP Aumasson2021-01-113-42/+72
|/
* Merge pull request #26 from jflopezfernandez/master•••Added some common IDE folders to gitignoreJean-Philippe Aumasson2019-12-151-0/+5
|\
| * Added some common IDE folders to gitignoreJose Fernando Lopez Fernandez2019-12-151-0/+5
|/
* Merge pull request #23 from altendky/return_for_test_failed•••Return non-zero when tests failJean-Philippe Aumasson2019-09-271-1/+4
|\
| * Initialize any_failed•••oops Kyle Altendorf2019-09-221-1/+1
| * Merge branch 'master' into return_for_test_failedKyle Altendorf2019-09-223-2/+7
| |\ | |/ |/|
* | Merge pull request #24 from altendky/separate_test_and_main•••Separate test function and main()Jean-Philippe Aumasson2019-09-223-2/+7
|\ \
| * | Separate test function and main()•••This allows for inclusion of the test function/file in a program with an existing main(). This is an isolated piece of #19. Kyle Altendorf2019-09-213-2/+7
|/ /
| * Return non-zero when tests fail•••This allows for shell scripts to check for success or failure without parsing the printed output. If the function name is later changed from main() it will also allow other code calling the tests to also easily check the pass or fail status. Kyle Altendorf2019-09-211-1/+4
|/
* Merge pull request #21 from altendky/full_siphash_format_specifiers•••Use size_t and fixed width format specifiers for printf()... for siphash.cJean-Philippe Aumasson2019-09-191-8/+5
|\
| * Correct \ alignmentKyle Altendorf2019-09-181-4/+4
| * Use size_t and fixed width format specifiers for printf()... for siphash.c•••It was forgotten in #17 to apply this to siphash.c as well as halfsiphash.c. Note that as-is this does change to a single 16-hex-char string for v0/1/2/3 in debug output rather than having 8 then a space then 8. If preferred it can be changed back to having the space and using `PRIx32` twice instead of `PRIx64` once. Kyle Altendorf2019-09-181-8/+5
|/
* Merge pull request #17 from altendky/format_specifiers•••Use size_t and fixed width format specifiers for printf()Jean-Philippe Aumasson2019-09-141-4/+5
|\
| * Use size_t and fixed with format specifiers for printf()Kyle Altendorf2019-09-131-4/+5
generated by cgit v1.2.3 (git 2.52.0) at 2026-04-11 05:17:08 +0000