summaryrefslogtreecommitdiff
path: root/open_issues/virtualization/fakeroot.mdwn
diff options
context:
space:
mode:
Diffstat (limited to 'open_issues/virtualization/fakeroot.mdwn')
-rw-r--r--open_issues/virtualization/fakeroot.mdwn43
1 files changed, 43 insertions, 0 deletions
diff --git a/open_issues/virtualization/fakeroot.mdwn b/open_issues/virtualization/fakeroot.mdwn
index f473977..f9dd475 100644
--- a/open_issues/virtualization/fakeroot.mdwn
+++ b/open_issues/virtualization/fakeroot.mdwn
@@ -22,3 +22,46 @@ License|/fdl]]."]]"""]]
<youpi> btw, I believe our fakeroot-hurd is close to working actually
<youpi> it's just a argv[0] issue supposed to be fixed by exec_file_name
but apparently not fixed in that case, for some reason
+
+
+## IRC, freenode, #hurd, 2013-08-26
+
+ < teythoon> also I looked into the fakeroot issue, aiui the problem is that
+ scripts are not handled correctly, right?
+ < teythoon> the exec server fails to locate the scripts file name, and so
+ it hands the file_t to the interpreter process and passes /dev/fds/3 as
+ script name
+ < teythoon> afaics that breaks e.g. python
+ < youpi> yes
+ < youpi> pinotree's exec_file_name is supposed to fix that, but for some
+ reason it doesn't work here
+ < pinotree> it was pochu's, not mine
+ < youpi> ah, right
+ < teythoon> ah I see, I was wondering about that
+ < pochu> it was working for a long time, wasn't it?
+ < pochu> and only stopped working recently
+ < youpi> did it completely stop?
+ < youpi> I have indeed seen odd issues
+ < youpi> I haven't actually checked whether it has completely stopped
+ working
+ < youpi> probably worth looking there first
+ < pinotree> gtk+3.0 fails, but other stuff like glib2.0 and gtester-using
+ stuff works
+ < teythoon> huh? I created tests like "#!/bin/sh\necho $0" and that says
+ /dev/fd..., and a python script doing the same doesn't even run, so how
+ can it work for a package build?
+ < youpi> it works for me in plain bash
+ < youpi> #!/bin/sh
+ < youpi> echo $0
+ < youpi> € $PWD/test.sh
+ < youpi> /home/samy/test.sh
+ < teythoon> it does !?
+ < youpi> yes
+ < youpi> not in fakeroot-hurd however, as we said
+ < teythoon> well, obviously it works when not being run under
+ fakeroot-hurd, yes
+ < youpi> ok, so we weren't talking about the same thing
+ < youpi> a mere shell script doesn't work in fakeroot-hurd indeed
+ < youpi> that's why we still use fakeroot-sysv
+ < teythoon> right
+ < youpi> err, -tcp