[Ltp-cvs] ltp ChangeLog,1.87,1.88

[mreed <mr...@us...>]

Update of /cvsroot/ltp/ltp
In directory sc8-pr-cvs6.sourceforge.net:/tmp/cvs-serv30072

Modified Files:
	ChangeLog 
Log Message:
Changes for the July release of LTP


Index: ChangeLog
===================================================================
RCS file: /cvsroot/ltp/ltp/ChangeLog,v
retrieving revision 1.87
retrieving revision 1.88
diff -C2 -d -r1.87 -r1.88
*** ChangeLog	10 Jul 2006 06:27:13 -0000	1.87
--- ChangeLog	17 Jul 2006 21:16:57 -0000	1.88
***************
*** 1,2 ****
--- 1,57 ----
+ LTP-20060717
+  - The tarball default-tests.tar.gz is a replacement for
+    testcases/pounder21/default-tests.tar.gz.  This new pounder config
+    enables the magic sysrq key when pounder starts.   
+  - A patch submitted by Derek Wong to reduce the memory requirements of
+    pounder's ramsnake test.
+  - A patch submitted by Will Newton that allows for compatibility changes
+    gcc 2.95.2 in th following files:
+      lib.c, lib64.c, test.c, test64.c, test_func.c, test_func64.c, tools.c
+  - Fixed ColdFire Makefile mistake in the syscall and syscalls/mmap directory
+  - Added a note for uClinux users in the top level Makefile
+  - A fix for failures in fcntl27 and fcntl28 for bugs 21614 and 23235.
+  - A fix submitted for make_tree.c by Jacky Malcles that fixes this
+    testcase by setting envp
+   - A fix submitted to Jacky Malcles that fixes read_checkzero.c.  The lseek
+    function allows the file offset to be set beyond the end of the existing
+    end-of-file of the file.  If data is later written at this point,
+    subsequent reads of the data is in the gap returns bytes of zeros until
+    data is actually written into the gap.
+  - In the testcase semget05.c the value of MAXIDS was changed for the specific
+    machine by reading the system limit for SEMMNI - The maximum number of
+    semaphore sets. This is a fix for bug 24745
+  - A fix submitted by Amit K. Amora that initializes the alarm received
+    code and allows the test to pass more than just once on 2.6.17-rc6
+    alarm05.c,
+  - A fix was submitted by Andy Echols for pan.c to fix an infinite
+    loop problem that occurs  in pan if runltp tries to run a test
+    that isn't present.
+  - A fix was submitted to cast TEST_RETURN to gid_t to avoid implicit casts
+    which tend to cause problems with the testcase  setregid03.c,
+  - A patch submitted by Jacky Malcles that fixes the problem where
+    i0_getevents() return value is not checked and may return 0 if
+    no events are available and may generate a SIGSEGV in the testcase
+    aiodio_append.c,
+  - Backed out the _USC_LIB change for several Makefiles because it was
+    breaking on the PowerPc platform on Fedora Core
+  - Added code to ignore looking for PID_MAX on powerpc, s390, and i386 to
+    fix build problems on newer kernel versions on the following files:
+     sysctl05.c, setpriority04.c sysfs01.c, sysfs02.c, sysfs03.c, sysfs04.c,
+     sysfs06. getdents01.c, sysctl03.c getsid02.c, sysctl01.c,wait402.c
+  - TCP.c was changed to delete broken whitespace and also the call for
+    accept(2) takes a socklen_t, not an int
+  - Changes were added to the following files to use memset() instead of
+    bzero():
+    tlibio.c, write_log.c, doio.c, iogen.c, fsstress.c, fsx-linux.c,
+    pthcli.c, pthserv.c, pth_str01.c, pth_str03.c, recvmsg01.c, sendmsg01.c,
+    crash01.c, crash02.c, pingpong6.c,test_getname.c,fancy_timed_loop.c, 
+    infinite_loop.c, run.c, timed_loop.c, snake.c, rpc1.c,pipeio.c, mc_recv.c
+  - Changes were added to the following files to use memcpy() instead of
+     bcopy():
+     serverCommunication.cpp, member.c, rpc1.c,pipeio.c,mc_recv.c
+ -  A series of  patches created by Mitsuru Chinen that  created some addtional 
+    network stress tests. 
+ -  A change submitted by Liang Shuang that fixes su01_su for the arm 
+    architecture
  LTP-20060615
  -Added a patch from Darrick Wong to reduce the memory requirements of