[scarthgab] commit da15df26e62bf9b030ec93d94d0182f4d9562e32

[san-skov-dk]

The CVE-2026-43895 for jq breaks compilation no matter building it native too.

src/.libs/bytecode.o
| cc1: fatal error: src/parser.c: No such file or directory
| compilation terminated.
| make[2]: *** [Makefile:1051: src/parser.lo] Error 1
| make[2]: *** Waiting for unfinished jobs....
| ../jq-1.7.1/src/jv_parse.c: In function 'found_string':
| ../jq-1.7.1/src/jv_parse.c:496:13: warning: comparison is always true due to limited range of data type [-Wtype-limits]
| 496 | if (c >= 0 && c <= 0x001f)
| |

[ernstp]

Commit da15df2 modifies both parser.c and parser.y

parser.c can be used as-is or regenerated from parser.y every time (with flex and bison). This is controlled by the maintainer-mode option.

However this error doesn't happen all the time for some reason. I can reproduce it on one computer but not on several others.

ping @anujm1

[san-skov-dk]

I use AlmaLinux 10 with AMD cpus as build host . Dont know if it has something to do with this? Not official supported.

[ernstp]

Perhaps you could attach your full tmp/work/XX/jq/1.7.1/temp/log.do_configure and tmp/work/XX/jq/1.7.1/temp/log.do_compile

[san-skov-dk]

log.do_compile.txt
log.do_configure.txt

[ernstp]

This seems to happen when parser.y has a newer mtime than parser.c . Which is caused by the patch process itself.

Make considers parser.c out of date, but since maintainer mode is not enabled it goes to "NOT building parser.c!", seen in the beginning of the compile log.