Commit report for master branch
New activity for FreeRADIUS (the high performance and highly configurable RADIUS server) ====== Extra dependencies for building debian packages Arran Cudbard-Bell@2013-03-21T22:32:09Z Files modified: * .travis.yml Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/ae2ef45ef5ecbd867ff76... ====== Fix travis-test Arran Cudbard-Bell@2013-03-21T22:25:06Z Files modified: * Makefile Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/a452a4f8c83b0cdb430e8... ====== More travis tests Arran Cudbard-Bell@2013-03-21T19:41:10Z Files modified: * .travis.yml * Makefile Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/2e26fd088b038e6cedc0b... ====== Enabled WITH_VERIFY_PTR for developer builds Alan T. DeKok@2013-03-21T19:10:40Z Files modified: * configure * configure.in Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/7f4aba05525ea011b11b6... ====== Remove redundant code Arran Cudbard-Bell@2013-03-21T17:41:30Z Files modified: * src/lib/valuepair.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/89bb28d8be0053202363e... ====== Cleanup some logging stuff Arran Cudbard-Bell@2013-03-21T17:35:53Z Files modified: * src/main/log.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/fa11fce44fd6d8468fe05... ====== Fix minor memory leak on detach Alan T. DeKok@2013-03-21T16:50:30Z Files modified: * src/modules/rlm_passwd/rlm_passwd.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/f0a6efc2c3141202ec418... ====== Added verification of VALUE_PAIR pointers via talloc Alan T. DeKok@2013-03-21T16:46:17Z Files modified: * src/lib/valuepair.c * src/main/evaluate.c * src/main/radiusd.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/08a7fd1c39149a854e219... ====== Remove unused / old files Alan T. DeKok@2013-03-21T14:54:22Z Files modified: * debian/freeradius.examples Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/8c2cc677eb02a05db8272... ====== Remove duplicate and unused package Die, libltdl, die. Alan T. DeKok@2013-03-21T14:46:59Z Files modified: * debian/control Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/94cfd7f99062949bcff71... ====== Remove references to rlm_dbm Alan T. DeKok@2013-03-21T14:46:30Z Files modified: * debian/control * debian/freeradius-utils.install Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/62de6d540aca185a1c790... ====== Merge pull request #239 from netflash/master Building debian package. Fixed some errors Alan DeKok@2013-03-21T14:45:30Z Files modified: * debian/control Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/3b0d518ecf5d5c6ea22a2... ====== Fixed debian/control build-dep section added 3 libraries Alexandr Romanov@2013-03-21T11:28:14Z Files modified: * debian/control Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/40cdae1c56256d195248e... ====== More debugging Arran Cudbard-Bell@2013-03-21T08:19:08Z Files modified: * src/main/log.c * src/main/radiusd.c * src/main/valuepair.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/ae425264deb22b54153b6... ====== See what talloc report gives us Arran Cudbard-Bell@2013-03-21T07:53:34Z Files modified: * src/main/valuepair.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/c7f2f7326dfc006e056a3... ====== We don't know it's not valid, just that it's not set Arran Cudbard-Bell@2013-03-21T07:45:27Z Files modified: * src/lib/valuepair.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/fd39bb5c9295071c46e86... ====== Add more opportunistic debugging in pairbasicfree Arran Cudbard-Bell@2013-03-21T07:43:09Z Files modified: * src/lib/valuepair.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/374155399dc65d4ad4aba... ====== Add assert to pairbasicfree to check for da Arran Cudbard-Bell@2013-03-21T07:29:58Z Files modified: * src/lib/valuepair.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/90fc55484512bc7c562d3... ====== Verify VPs are still viable in pair_list functions Arran Cudbard-Bell@2013-03-21T07:21:47Z Files modified: * src/main/valuepair.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/ad0c8bcb1465a7cfeac4e... ====== Minor typo Arran Cudbard-Bell@2013-03-21T06:27:26Z Files modified: * src/modules/rlm_eap/eap.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/d217abd2d7df9f435ad82... ====== Add assert to catch NULL source context Arran Cudbard-Bell@2013-03-21T06:27:26Z Files modified: * src/modules/rlm_krb5/rlm_krb5.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/f746366c11354fba59c3b... ====== Don't segv if we didn't managed to load one of the eap methods Arran Cudbard-Bell@2013-03-21T05:55:28Z Files modified: * src/modules/rlm_eap/eap.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/e06a68eedcff9e59b9a09... ====== Kill another smurf Arran Cudbard-Bell@2013-03-21T05:54:59Z Files modified: * src/modules/rlm_eap/eap.c * src/modules/rlm_eap/rlm_eap.h Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/08a70c18986549c9d905d... ====== Apparently none of the range constructs work in Travis Arran Cudbard-Bell@2013-03-21T05:33:07Z Files modified: * src/tests/Makefile Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/e567d5ce3c0ec03e13d5f... ====== Add -P argument to always write PID, and fix tests Arran Cudbard-Bell@2013-03-21T05:28:09Z Files modified: * src/main/radiusd.c * src/tests/Makefile * src/tests/gdb.conf Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/25bde7a2b7240045674b7... ====== GDB is apparently not available by default Arran Cudbard-Bell@2013-03-21T02:52:14Z Files modified: * .travis.yml Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/bd9136f51044f0ee99be7... ====== Attempt at autobacktrace on segv Arran Cudbard-Bell@2013-03-21T00:32:04Z Files modified: * src/tests/Makefile * src/tests/gdb.conf Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/fb6498cc489b86221293c... ====== Fix typo. Set "from" to NULL after we steal all of the VPs Alan T. DeKok@2013-03-21T00:22:11Z Files modified: * src/lib/valuepair.c Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/5442305f3876d20aaf0e8... ====== -- This commit summary was generated @2013-03-22T00:00:01Z by lgfeed version 0.00 (https://github.com/arr2036/lgfeed).
On Fri, Mar 22, 2013 at 12:00:01AM +0100, The git bot wrote:
====== Remove redundant code
Arran Cudbard-Bell@2013-03-21T17:41:30Z Files modified: * src/lib/valuepair.c
Commit diff: https://github.com/FreeRADIUS/freeradius-server/commit/89bb28d8be0053202363e... ======
Observation: the code now does #define VERIFY(_x) _x=talloc_get_type(_x, VALUE_PAIR) and my understanding is that if the type is wrong, _x will be set to null. http://talloc.samba.org/talloc/doc/html/group__talloc.html#ga1ee43e9ef59fc4e... "Returns: The properly casted pointer given by ptr, NULL on error." However the code then goes ahead and indirects through vp without first checking if vp is null, and there are lots of instances of this. Is that intentional? I guess if all you want is a segfault on type error then it's OK (but in that case I don't see why you don't use talloc_get_type_abort) Regards, Brian. --- a/src/lib/valuepair.c +++ b/src/lib/valuepair.c @@ -41,7 +41,7 @@ RCSID("$Id$") /* * Requires typeof(), which is in most modern C compilers. */ -#define VERIFY(_x) _x=talloc_get_type_abort(_x, VALUE_PAIR) +#define VERIFY(_x) _x=talloc_get_type(_x, VALUE_PAIR) #else #define VERIFY(_x) #endif @@ -122,18 +122,10 @@ void pairbasicfree(VALUE_PAIR *vp) VERIFY(vp); /* - * The lack of DA probably means something has gone - * wrong, try and get as much info as we can before - * calling talloc_free (at which point we'll probably get - * an abort). + * The lack of DA means something has gone wrong */ if (!vp->da) { fr_strerror_printf("VALUE_PAIR has NULL DICT_ATTR pointer (probab - - if (!talloc_get_type(vp, VALUE_PAIR)) { - fr_strerror_printf("Pointer being freed is NOT a VALUE_PA - talloc_get_name(vp)); - }
Brian Candler wrote:
#define VERIFY(_x) _x=talloc_get_type(_x, VALUE_PAIR)
and my understanding is that if the type is wrong, _x will be set to null.
Yes.
However the code then goes ahead and indirects through vp without first checking if vp is null, and there are lots of instances of this.
Is that intentional?
No. My $0.02 is that if something goes that wrong, the only thing you can do is to exit(). Arran wants to print out information about the "wrong" context. But if memory has been corrupted, all bets are off.
I guess if all you want is a segfault on type error then it's OK (but in that case I don't see why you don't use talloc_get_type_abort)
You would. It's what I had originally. The change is wrong. My thoughts are here: http://24.media.tumblr.com/tumblr_lqao3admCI1qzmcqxo1_400.gif Alan DeKok.
participants (3)
-
Alan DeKok -
announceļ¼ freeradius.org -
Brian Candler