Changes in xlint http://nxr.netbsd.org/rss/src/tests/usr.bin/xlint/ en Copyright 2005 Java lint: add more details to message about snprintf combinations Mon Apr 20 22:00:00 UTC 2026 rillig tests/lint: add test cases for possibly zero-extended sign bit Mon Apr 20 22:00:00 UTC 2026 rillig lint: warn about 'const' or 'volatile' in function return type Mon Apr 20 22:00:00 UTC 2026 rillig tests/lint: remove outdated comment, add volatile cast Mon Apr 20 21:00:00 UTC 2026 rillig lint: in C11 mode, allow typedef to redefine an existing type Sun Apr 19 16:00:00 UTC 2026 rillig tests/lint: increase test timeout for pmax<br/><br/>The lint1 test timed out after 300 seconds, reaching msg_333. 400<br/>seconds should be enough to reach all test cases. Wed Apr 15 05:00:00 UTC 2026 rillig lint: add query for negation of minimum signed value<br/><br/>This helped double-check that the remaining negations in the<br/>softfloat/bits64 code are fine. There are some false positives, due to<br/>int16 being typedef'd to int, thus being int32 realistically. Tue Apr 07 20:00:00 UTC 2026 rillig lint: do not warn about lossy cast from floating to integer<br/><br/>The check is intended to catch unintentional loss of accuracy. Having an<br/>explicit cast in the code makes the intention clear enough. Sat Mar 28 11:00:00 UTC 2026 rillig lint: allow more names for enum constants marking the maximum value Sat Mar 28 00:00:00 UTC 2026 rillig lint: in case of lossy floating point conversions, print exact value<br/><br/>When running lint on sparc64 with target sparc64, some tests print<br/>warnings in cases where they shouldn't. To track down their cause,<br/>print the exact floating point value in the message, in hexadecimal.<br/><br/>As the snprintf(3) manual page says, there are multiple equivalent<br/>representations, so just match them using the placeholder "....." in the<br/>"expect" comments. In this particular case, the message is missing<br/>completely, so the output of the failed test will reveal the actual<br/>value.<br/><br/>The affected test cases are:<br/>* msg_380.c:38, converting -9223372036854775809.0 to 'long long'<br/>* msg_380.c:39, converting -9223372036854775808.0 to 'long long'<br/>* msg_381.c:38, converting -9223372036854775809.0 to 'long long'<br/>* msg_381.c:39, converting -9223372036854775808.0 to 'long long' Fri Mar 27 21:00:00 UTC 2026 rillig lint: in _Generic expressions, only evaluate the matching branches Wed Mar 18 06:00:00 UTC 2026 rillig tests/lint: explain warning about an enum forward declaration Sun Mar 15 18:00:00 UTC 2026 rillig tests/lint: clean up comments, add a few test cases for snprintb Sun Mar 15 18:00:00 UTC 2026 rillig lint: in no-prototype functions, mention the function name Sun Mar 15 08:00:00 UTC 2026 rillig lint: for a non-scalar controlling expression, mention the type Sun Mar 15 05:00:00 UTC 2026 rillig lint: in a sizeof operator, do not check for division by zero Sun Mar 15 07:00:00 UTC 2026 rillig tests/lint: test reachability after a switch statement Sat Mar 14 20:00:00 UTC 2026 rillig lint: accept '$' in preprocessor macro expansions Wed Mar 11 00:00:00 UTC 2026 rillig tests/lint: test preprocessing tokens (lint1) and noreturn (lint2) Wed Mar 11 00:00:00 UTC 2026 rillig lint: support __float128<br/><br/>Needed by libquadmath. Tue Jan 20 23:00:00 UTC 2026 rillig