Re: Multiple patches for Kannel
Aris Adamantiadis <[email protected]>
| Newsgroups | gmane.comp.mobile.kannel.devel |
|---|---|
| Message-ID | <[email protected]> |
Hello, While the test haven't completed, I will not have much time later to work with the patches so here it is. As you can see, it has been split into 48 different files. (hence the .tgz) Not all of them are of equal importance. They all apply to kannel 1.5.0-release. I did not try to cherry pick them in the current branch. Some special patches: 0001-Changements-de-Martin.patch: This fixes a problem with the Oracle SQL expression (Already fixed in svn I think), but also a problem in text conversion, in which an SMS containing only ciphers would be converted into ascii (This causes problem when kannel is used to send one time passwords). 0047-cherry-pick-of-gsoap-ParlayX-support.patch: This is a backport of ParlayX on 1.5.0. 0048-Fix-out-of-bound-read-strcpy-to-memcpy.patch: A potential bug I fixed in the ParlayX code. I believe this will apply perfectly on the svn branch. It's possible that some patches contain /* flawfinder: ignore */. It is perfectly safe not to merge these lines ! Kind regards, Aris Le 5/02/13 17:16, Aris Adamantiadis a écrit : > Hi, > > When speaking with the rest of the team, we agreed that is was better to > wait until we finish another kannel-related task (back-porting the > ParlayX to our current patchset and testing it) before posting the > patches, in order not to miss anything. > You will hear from me before the end of the week. > Which format is the better ? A big .patch with all differences ? Is it > ok if I get that diff from kannel release 1.5.0 ? > > Kr, > > Aris > Le 5/02/13 14:18, spameden a écrit : >> Interesting find.. Would love to see what actually you've changed and >> what's considered to be insecure. I think you can post your diff to >> this devel list. >> >> Many thanks for your work. >> >> 2013/2/5 Aris Adamantiadis <[email protected]>: >>> Dear Kannel developers, >>> >>> During a security audit of Kannel, we identified several weaknesses in >>> the code, mostly unsafe C functions or data copying used without bound >>> checkings. These patches currently run in production on our site, but >>> we'd prefer to give them out to the community (and this makes our update >>> process easier as well). >>> >>> What is the best way to provide you with these patches ? Currently, they >>> are being tracked in a local git repository. I can make the work of >>> porting them to the latest subversion repository, but you would still >>> need someone to review and publish them on your svn. >>> >>> How can we proceed ? >>> >>> Kind regards, >>> >>> Aris Adamantiadis >>> >>> output of "git diff old_prod..new_prod --stat": >>> >>> addons/opensmppbox/gw/opensmppbox.c | 2 +- >>> gw/smsbox.c | 6 +- >>> gw/smsc/smsc.c | 2 +- >>> gw/smsc/smsc_at.c | 6 +- >>> gw/smsc/smsc_cgw.c | 2 +- >>> gw/smsc/smsc_cimd.c | 47 ++++++------ >>> gw/smsc/smsc_cimd2.c | 4 +- >>> gw/smsc/smsc_emi_x25.c | 74 +++++++++--------- >>> gw/smsc/smsc_ois.c | 140 >>> +++++++++++++++++------------------ >>> gw/smsc/smsc_sema.c | 66 ++++++++++------- >>> gw/smsc/smsc_sema.h | 2 +- >>> gw/smsc/smsc_soap.c | 27 ++++--- >>> gw/wap-appl.c | 10 ++- >>> gw/wap_push_ppg.c | 10 ++- >>> gwlib/accesslog.c | 6 +- >>> gwlib/conn.c | 2 +- >>> gwlib/date.c | 2 +- >>> gwlib/gw_uuid.c | 6 +- >>> gwlib/gwthread-pthread.c | 2 +- >>> gwlib/log.c | 33 +++++---- >>> gwlib/octstr.c | 4 +- >>> gwlib/utils.c | 13 ---- >>> gwlib/utils.h | 6 -- >>> test/fakewap.c | 8 +- >>> utils/run_kannel_box.c | 2 +- >>> utils/seewbmp.c | 8 +- >>> utils/start-stop-daemon.c | 26 ++++--- >>> wap/cookies.c | 8 +- >>> wap/wsp_session.c | 4 +- >>> wmlscript/wsstream_data.c | 12 +-- >>> wmlscript/wsstream_file.c | 6 +- >>> 31 files changed, 288 insertions(+), 258 deletions(-) >>> >> > >
patches.tgz
(application/x-gzip, 45.9 KB) - not displayed