1*1d3f2ddcSSimon J. Gerraty# $NetBSD: opt-debug-parse.mk,v 1.7 2022/02/09 21:09:24 rillig Exp $ 2956e45f6SSimon J. Gerraty# 3956e45f6SSimon J. Gerraty# Tests for the -dp command line option, which adds debug logging about 4956e45f6SSimon J. Gerraty# makefile parsing. 5956e45f6SSimon J. Gerraty 69f45a3c8SSimon J. Gerraty.MAKEFLAGS: -dp 79f45a3c8SSimon J. Gerraty 8956e45f6SSimon J. Gerraty# TODO: Implementation 9956e45f6SSimon J. Gerraty 109f45a3c8SSimon J. Gerraty# Before parse.c 1.639 from 2022-01-08, PrintStackTrace and other diagnostics 119f45a3c8SSimon J. Gerraty# printed a wrong line number, using the last line before the loop body, while 129f45a3c8SSimon J. Gerraty# it should rather be the line number where the .for loop starts. 139f45a3c8SSimon J. Gerraty# 149f45a3c8SSimon J. Gerraty# Before parse.c 1.643 from 2022-01-08, PrintStackTrace tried to be too clever 159f45a3c8SSimon J. Gerraty# by merging stack trace entries, printing confusing line numbers as a result. 169f45a3c8SSimon J. Gerraty.for \ 179f45a3c8SSimon J. Gerraty var \ 189f45a3c8SSimon J. Gerraty in \ 199f45a3c8SSimon J. Gerraty value 209f45a3c8SSimon J. Gerraty.info trace with multi-line .for loop head 219f45a3c8SSimon J. Gerraty.endfor 229f45a3c8SSimon J. Gerraty 23*1d3f2ddcSSimon J. Gerraty# Before parse.c 1.641 from 2022-01-08, the debug log said it returned to 249f45a3c8SSimon J. Gerraty# the line of the '.include' instead of the line following it. 259f45a3c8SSimon J. Gerraty.include "/dev/null" 269f45a3c8SSimon J. Gerraty 279f45a3c8SSimon J. Gerraty 289f45a3c8SSimon J. Gerraty# In .for loops with multiple variables, the variable details are included in 299f45a3c8SSimon J. Gerraty# the stack trace, just as with a single variable. 309f45a3c8SSimon J. Gerraty.for a b c in 1 2 3 ${:U4 5 6} 319f45a3c8SSimon J. Gerraty.info trace 329f45a3c8SSimon J. Gerraty.endfor 339f45a3c8SSimon J. Gerraty 349f45a3c8SSimon J. Gerraty 359f45a3c8SSimon J. Gerraty.MAKEFLAGS: -d0 369f45a3c8SSimon J. Gerraty 379f45a3c8SSimon J. Gerratyall: .PHONY 38