summaryrefslogtreecommitdiff
path: root/mk/check
AgeCommit message (Collapse)AuthorFilesLines
2008-01-16Make the default rpath for USE_CHECK_SHLIBS_ELF a platform variable.joerg1-2/+2
Define the values for NetBSD and DragonFly.
2007-11-30The interpreter check for packages which are their own interpreter,bjs1-2/+6
e.g. lang/perl5, was failing. Adding a check for ${DESTDIR}"$$interp" fixes this.
2007-11-28Print a warning for every file that has /bin/sh as interpreter, but onlyrillig1-1/+6
on Solaris.
2007-11-07Activated the check-interpreter for PKG_DEVELOPERs. I don't remember whyrillig1-1/+5
activating this has taken so long.
2007-10-13Fix the .include lines so that make looks in the right place first.dsl1-9/+9
Remember .include "foo.mk" is looked for (first) in the directory that contains the makefile being processed (like in C), so remove all the ${.PARSEDIR} and ../ sequences that just cause grief.
2007-09-27Even though VARBASE is sometimes below LOCALBASE, some users want torillig1-1/+3
check it. Don't prevent that in CHECK_FILES_STRICT=yes mode.
2007-09-13In bulk builds with a custom PKGSRCDIR, that directory should not appearrillig1-1/+3
in binary packages. This check can be enabled by adding "pkgsrc" to CHECK_WRKREF.
2007-09-12check-files shouldn't complain about files in VARBASE, since they cannotrillig1-1/+4
be represented in the PLIST.
2007-09-05Now it's possible to add custom directories to the check for directoryrillig1-1/+9
references.
2007-08-21Allow the existence of symlinks that point nowhere.jlam1-3/+3
2007-08-21Add actual license and a small description of what this is doing.joerg1-0/+40
2007-08-20Add a cross-compile and DESTDIR safe alternative to ldd forjoerg3-2/+142
check-shlibs. This allows more stricter detection (e.g. it checks whether the package links against a build dependency). It also checks whether the rpath contains wrkdir references (like qt3-libs). It depends on readelf and is only available for ELF systems for now. To test it, add USE_CHECK_SHLIBS_ELF to mk.conf. Supported-by: Google SoC 2007
2007-08-14Fix a number of bugs in the DESTDIR support:joerg1-2/+2
- check-perms.mk was not DESTDIR aware, prefix files before passing it to the directory extraction - PKG_FILELIST_CMD was calling pkg_info, which is fine for normal installation, but fails of course for DESTDIR. Just drop the @ lines from ${_DEPENDS_PLIST} and use that. - To make he former work, ensure that _flavor-generate-metadata is part of _INSTALL_ALL_TARGETS. It was normally a dependency of _flavor-register, but that is skipped for DESTDIR. - Remove ${_DEPENDS_PLIST} when running install-clean.
2007-08-13Prevented the files check from being confused by concurrent instances ofrillig1-1/+3
"make replace". Since these are currently not locked properly, they may remove files from PKG_DBDIR while another package is being installed.
2007-07-23When using check-perms, also depend on sysutils/checkperms.joerg1-1/+3
2007-04-14Teach pkgsrc about HP-UX.tnn1-2/+2
2007-03-24Made this check work again. The regular expression did not match relativeheinz1-6/+7
paths from .PLIST. Changed the working directory to ${DESTDIR}${PREFIX} so files can be accessed correctly. Extended the relative paths to absolute paths again for better clarity. Improved (IMHO) some wording.
2007-03-24More DESTDIR fixes in warnng/error messagesheinz1-5/+5
2007-03-24Removed unnecessary shell code:heinz1-2/+2
p|p) is the same as p)
2007-03-20Fix check-interpreter when using USE_DESTDIR. From Klaus Heinz onjoerg1-3/+3
tech-pkg.
2007-03-18Typo.rillig1-2/+2
2007-03-16Added the variable declarations for the "show-all" target.rillig5-5/+25
2007-03-16Don't check files in ${PACKAGES} and ${DISTDIR}.rillig1-1/+5
2007-03-15For all those who are interested in the inner workings of pkgsrc, thererillig1-1/+5
is a new target "show-all" that fits to the existing "debug", "show-tools", "show-vars" targets. It prints a list of the variables that make up the public interface to pkgsrc. Running this target is especially useful if you want to do some things, you know that they must have been implemented but you don't know what it is called. It also shows the "class" of a variable (user-defined, package-defined, system-defined).
2007-03-15The CHECK_WRKREF variable has changed type: It is now a list ofrillig1-16/+18
directory aliases. That makes it possible to check for references into TOOLS_DIR and WRKSRC, but to leave out the other infrastructure directories. On IRIX, x11/gtk2 failed to build because of an unknown library -lharfbuzz. That library turned out to be an internal dependency of graphics/cairo, which had made it into the installed libpangocairo.la file. With this change, that bug would have been caught earlier.
2007-03-09Renamed some more flavor-specific targets to have the form _flavor-*.rillig1-12/+17
Split check-vulnerable into a general an a flavor-specific part, for consistence with all the other targets.
2007-03-08Documented how to use checkperms in bulk builds.rillig1-1/+5
2007-03-02Remove support for obsolete and unused NO_PKG_REGISTER.wiz1-16/+1
2007-01-07Fixed the mismatch between absolute and relative paths.rillig1-4/+4
2007-01-06Moved the code that installs check-files on the hook from install.mk torillig1-1/+5
check-files.mk, for two reasons: 1. It belongs there. 2. Now check-files is run before check-perms and check-interpreter.
2007-01-06Reformatted the help text to be more human-readable instead ofrillig1-7/+15
pkglint-readable.
2007-01-02Removed the need to define the PKGSRCDIR environment variable.rillig5-18/+20
2006-12-31Made the portability check more portable by not using a hard-codedrillig2-6/+6
"/dev/stderr" in the AWK programs. That device file doesn't exist on IRIX, and it isn't required by POSIX either.
2006-12-15Remove trailing tabs.martti1-2/+2
2006-12-12Made all .include lines look the same.rillig1-7/+7
2006-12-12On Solaris 5.9, the ksh cannot handle null bytes in the input. It'srillig1-2/+13
documented in a SunSolve document, but that document is not accessible to the public. There's no better way than to start a subshell executing sed(1) in this case. This makes the test even slower on these machines, and is not even guaranteed to work in all cases. That's life.
2006-11-17On Interix, the -c option of checkperms is disabled, as checkperms doesrillig1-2/+6
not know about Interix yet. This is a temporary work-around for PR 34968.
2006-11-13Instead of $0 and $1 .. $NF, the function check_test_eqeq should testrillig1-5/+6
the string in the ''line'' variable. This will make koffice build again.
2006-11-12Added a missing line continuation in the code for CHECK_FILES_STRICT.rillig1-2/+2
Apparently no one has been using that code since June 2006.
2006-11-12Moved the test whether this check should be enabled to this file.rillig1-13/+7
Removed the dependency to NO_PKG_REGISTER. Renamed the check-interpreter target to _check-interpreter.
2006-11-12Added the CHECK_PERMS_AUTOFIX variable, which allows packages to sayrillig1-8/+17
that the file permissions should be fixed automatically. Moved the test whether this check should be enabled from ../install/install.mk to here. Renamed the target from check-perms to _check-perms.
2006-11-12Improved the formatting of the comment and made the code a littlerillig1-21/+22
shorter.
2006-11-12Moved the check for whether check-wrkref should be run fromrillig1-2/+6
install/install.mk to check/check-wrkref.mk. Renamed the check-wrkref target to _check-wrkref.
2006-11-12Improved the readability of the header comment.rillig1-24/+25
CHECK_WRKREF may be set to "wrkobjdir" to check for references to WRKOBJDIR. Added some error checking. Removed the dependency on NO_PKG_REGISTER.
2006-11-11Fixed the comment.rillig1-2/+2
2006-11-11Moved the check whether check-shlibs should be run fromrillig1-2/+7
install/install.mk to check/check-shlibs.mk. Renamed check-shlibs to _check-shlibs. Since it had been declared as "private" before, this will not break anything.
2006-11-11Added the usual header comment.rillig1-22/+26
Removed the special case handling of NO_PKG_REGISTER. What does checking for shared libraries have to do with whether the package is registered in the package database or not? Removed the check on CHECK_SHLIBS_SUPPORTED, which will be reintroduced soon.
2006-11-11Improved a comment.rillig1-3/+3
2006-11-10Fixed the warning from GNU awk that in strings, \$ is the same as arillig1-2/+2
plain $.
2006-11-10Refined the pattern for lines containing bad macros. White-space isrillig1-2/+2
allowed between the "#" and the "define", and the value must be of the form "${varname}/...", that is, starting with a double quote, followed by "${" and a variable name, and directly behind the closing brace must be a slash. This should catch most false positives while still being useful.
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-07 16:09:58 +0000