Closed Olf0 closed 2 years ago
Thanks or review, analysis and fixing MR, merged.
You are right, that variable was supposed to be an env var, or be set for either testing of whatever reason a user may have to turn this off. (It may have been added before there was the UI option to disable that too...)
Which leaves open the nonexisting documentation - what would be the best place to put that? As far as I'm aware we don't have an "innards of PM" section of documentation at the moment.
Which leaves open the non-existing documentation - what would be the best place to put that?
The PM-wiki here at GH.
As far as I'm aware we don't have an "innards of PM" section of documentation at the moment.
~~Correct. New sections are easy to create, many of the extant ones are just stubs, others contain some collected information which needs to be overhauled. Thus creating proper content is the hard but crucial part. Feel free to draft a new section; I may try as well but have enough open tasks, that this will be mid-January, earliest. At least I now know what to document; thanks for that information.~~ → I did not look properly, see next message.
But we do already have a section for environment variables, see https://github.com/sailfishos-patches/patchmanager/wiki/Environment-variables IMO it fits well there.
Documented.
BTW, it is actually used in https://github.com/sailfishos-patches/patchmanager/blob/master/src/bin/patchmanager-daemon/patchmanagerobject.cpp#L1822 to pass the UI configuration to disable bitness mangling to the script.
While reviewing the shell script
pm_apply
, specifically the new functionality of mangling library paths, I wondered how the disabling of the whole mangling is supposed to work.a. In line 27 the variable
DISABLE_MANGLING
is evaluated for being zero length, but never set (at any place I am aware of); hence I assume it is meant to be set as an environment variable. Q: Is this documented somewhere?b. If
DISABLE_MANGLING
contains something or sourcing the file/etc/patchmanager/manglelist.conf
does not set the variableMANGLE_CANDIDATES
to something, this variable stays empty (i.e., initialised, but zero length). But then thefor
loop in the functionmangle_libpath
is entered once with an empty string forp
, consequentlytotl_lines
=cand_lines
= n (with n ≥ 1), … IMO this cannot yield anything right.Hence I suggest to prefix the call of the function
mangle_libpath
(in the script body) with[ -n "$MANGLE_CANDIDATES" ] &&
per MR #219.