Source-Changes-HG archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

[src/trunk]: src/usr.bin/make/unit-tests make: document the bad state in whic...



details:   https://anonhg.NetBSD.org/src/rev/daa1e9adb5fc
branches:  trunk
changeset: 952773:daa1e9adb5fc
user:      rillig <rillig%NetBSD.org@localhost>
date:      Tue Feb 16 19:43:09 2021 +0000

description:
make: document the bad state in which the test varcmd.mk is

diffstat:

 usr.bin/make/unit-tests/varcmd.mk |  14 ++++++++++++--
 1 files changed, 12 insertions(+), 2 deletions(-)

diffs (27 lines):

diff -r 117735c2c968 -r daa1e9adb5fc usr.bin/make/unit-tests/varcmd.mk
--- a/usr.bin/make/unit-tests/varcmd.mk Tue Feb 16 19:01:18 2021 +0000
+++ b/usr.bin/make/unit-tests/varcmd.mk Tue Feb 16 19:43:09 2021 +0000
@@ -1,6 +1,17 @@
-# $NetBSD: varcmd.mk,v 1.5 2020/10/24 08:50:17 rillig Exp $
+# $NetBSD: varcmd.mk,v 1.6 2021/02/16 19:43:09 rillig Exp $
 #
 # Test behaviour of recursive make and vars set on command line.
+#
+# FIXME: The purpose of this test is unclear.  The test uses six levels of
+# sub-makes, which makes it incredibly hard to understand.  There must be at
+# least an introductory explanation about what _should_ happen here.
+# The variable names are terrible, as well as their values.
+#
+# This test produces different results if the large block with the condition
+# "scope == SCOPE_GLOBAL" in Var_SetWithFlags is removed.  This test should
+# be rewritten to make it clear why there is a difference and why this is
+# actually intended.  Removing that large block of code makes only this test
+# and vardebug.mk fail, which is not enough.
 
 FU=    fu
 FOO?=  foo
@@ -57,4 +68,3 @@
 
 six:   show-v
        @${.MAKE} -f ${MAKEFILE} V=override show-v
-



Home | Main Index | Thread Index | Old Index