Re: [OpenWrt-Devel] [PATCH procd] move /dev/shm to /tmp/shm

2015-09-16 Thread John Crispin
On 15/09/2015 17:05, Daniel Gimpelevich wrote: > I see procd got an update. Any reason this patch was left out? > > On Fri, 2015-09-11 at 13:05 -0700, Daniel Gimpelevich wrote: >> Is there an ETA for merging this? >> >> On Sun, 2015-07-19 at 19:14 -0700, Daniel Gimpelevich wrote: >>> I'm

Re: [OpenWrt-Devel] [PATCH procd] move /dev/shm to /tmp/shm

2015-09-15 Thread Daniel Gimpelevich
I see procd got an update. Any reason this patch was left out? On Fri, 2015-09-11 at 13:05 -0700, Daniel Gimpelevich wrote: > Is there an ETA for merging this? > > On Sun, 2015-07-19 at 19:14 -0700, Daniel Gimpelevich wrote: > > I'm intending for this fix to be backported to CC as well. > > > >

Re: [OpenWrt-Devel] [PATCH procd] move /dev/shm to /tmp/shm

2015-09-11 Thread Daniel Gimpelevich
Is there an ETA for merging this? On Sun, 2015-07-19 at 19:14 -0700, Daniel Gimpelevich wrote: > I'm intending for this fix to be backported to CC as well. > > On Tue, 2015-07-14 at 09:59 -0700, Daniel Gimpelevich wrote: > > OpenSSL session caches create shared memory segments that appear as > >

Re: [OpenWrt-Devel] [PATCH procd] move /dev/shm to /tmp/shm

2015-07-19 Thread Daniel Gimpelevich
I'm intending for this fix to be backported to CC as well. On Tue, 2015-07-14 at 09:59 -0700, Daniel Gimpelevich wrote: OpenSSL session caches create shared memory segments that appear as files under /dev/shm. If there is not enough room there, execution terminates with SIGBUS upon access.

Re: [OpenWrt-Devel] [PATCH procd] move /dev/shm to /tmp/shm

2015-07-14 Thread John Crispin
what kind of usage do you expect ? On 11/07/2015 03:58, Daniel Gimpelevich wrote: Since the /dev filesystem is tiny, /dev/shm needs to live somewhere else. Signed-off-by: Daniel Gimpelevich dan...@gimpelevich.san-francisco.ca.us --- a/initd/early.c 2015-06-19 14:01:25.0 -0700 +++

Re: [OpenWrt-Devel] [PATCH procd] move /dev/shm to /tmp/shm

2015-07-14 Thread Daniel Gimpelevich
OpenSSL session caches create shared memory segments that appear as files under /dev/shm. If there is not enough room there, execution terminates with SIGBUS upon access. Even a single instance would exceed 512K. On Tue, 2015-07-14 at 11:50 +0200, John Crispin wrote: what kind of usage do you

[OpenWrt-Devel] [PATCH procd] move /dev/shm to /tmp/shm

2015-07-10 Thread Daniel Gimpelevich
Since the /dev filesystem is tiny, /dev/shm needs to live somewhere else. Signed-off-by: Daniel Gimpelevich dan...@gimpelevich.san-francisco.ca.us --- a/initd/early.c 2015-06-19 14:01:25.0 -0700 +++ b/initd/early.c 2015-07-10 15:04:23.270143065 -0700 @@ -66,16 +66,20 @@