diff options
Diffstat (limited to 'hurd/translator/tmpfs/discussion.mdwn')
| -rw-r--r-- | hurd/translator/tmpfs/discussion.mdwn | 73 |
1 files changed, 72 insertions, 1 deletions
diff --git a/hurd/translator/tmpfs/discussion.mdwn b/hurd/translator/tmpfs/discussion.mdwn index 20aba837..72400121 100644 --- a/hurd/translator/tmpfs/discussion.mdwn +++ b/hurd/translator/tmpfs/discussion.mdwn @@ -1,4 +1,4 @@ -[[!meta copyright="Copyright © 2011, 2012, 2013 Free Software Foundation, +[[!meta copyright="Copyright © 2011, 2012, 2013, 2014 Free Software Foundation, Inc."]] [[!meta license="""[[!toggle id="license" text="GFDL 1.2+"]][[!toggleable @@ -19,6 +19,8 @@ License|/fdl]]."]]"""]] * [[!GNU_Savannah_bug 26751]] +[[!toc]] + # [[Maksym_Planeta]] @@ -430,3 +432,72 @@ License|/fdl]]."]]"""]] <youpi> ok <youpi> but that indeed means writeback of ext2fs works, which is a good sign :) + + +# IRC, freenode, #hurd, 2013-10-04 + + <teythoon> btw, I noticed that fifos do not work on tmpfs + <braunr> teythoon: tmpfs seems limited, yes + <teythoon> that's annoying b/c /run is a tmpfs on Debian and sysvinit + creates a crontrol fifo there + <teythoon> I wonder why I didn't notice that before + <braunr> also, fifos, like symlinks, can be shortcircuited in libdiskfs + <braunr> i wonder if that has anything to do with the problem at hand + +[[mtab/discussion]], *Multiple mtab Translators Spawned*. + + <teythoon> b/c this breaks reboot & friends + <teythoon> I do too + <teythoon> b/c I cannot find any shortcut related code in tmpfs + <braunr> well, it's optional normally + <braunr> so that's ok + <braunr> but has it really been tested when the option wasn't there ? :) + <teythoon> yes, but the tmpfs requests this by setting diskfs_shortcut_fifo + = 1; + <pinotree> hm i remember tmpfs was said to be working with + sockets/fifos/etc, back then when it was fixed + <braunr> teythoon: oh + + +## IRC, freenode, #hurd, 2013-10-11 + + <teythoon> this will have to wait for the next hurd pkg unfortunately, b/c + I broke tmpfs by accident :-/ + <pinotree> how so? + <teythoon> the dropping of privileges broke passive translators and mkfifo + <braunr> there actually is a reason why those are run as root or with the + privilege of their owner + <braunr> privileges should be decoupled from identity + <teythoon> yes + + +## IRC, freenode, #hurd, 2013-11-08 + + <teythoon> braunr: I'm investigating this port leak of mine + <teythoon> well, I thought I introduced one + <teythoon> but I'm not too sure anymore + <teythoon> the setting is this + <teythoon> i start an active tmpfs translator, bind it to foo + <teythoon> then, i create foo/bar with a passive translator entry + <teythoon> i access foo/bar, the passive translator is started + <teythoon> my test suite now covers several methods of making that + translator go away + <teythoon> killing it with 15 or 9 is fine, i.e. does not make the first + tmpfs leak ports + <teythoon> however, doing settrans -g foo/bar does for some reason + <teythoon> i think my code is fine, i spent considerable time on tracking + down this problem, always thinking that i must have introduced it + <teythoon> but another thing just cought my eye, the first tmpfs translator + says this when i do settrans -g foo/bar: + <teythoon> tmpfs/tmpfs: pthread_create: Resource temporarily unavailable + <teythoon> could it be that a no-sender notification is ignored b/c the + handler thread failed to start ? + <braunr> teythoon: i saw this pthread error too + + +# IRC, freenode, #hurd, 2014-02-09 + + <gg0> remounting tmpfs doesn't work if in use + http://paste.debian.net/plain/80937/ + <gg0> you will also get a pthread_create: Resource temporarily unavailable + <youpi> iirc the pthread_create warning happens for any kind of translator |