New in This Release
New in Purify 5.0.1
New in Purify 5.0
New in Purify 4.4
New in Purify 4.3
In addition to support for code generated by the Apex 3.0.0 C++ and Ada compilers, this release provides Apex GUI integration in the form of the Purify Viewer Edit, Debug, Check-in, Check-out, and JIT debugging features.
New In Purify 4.2
New In Purify 4.1
This release supports HP-UX 10.30, including kernel threads.
This release supports PA-RISC 2.0 object files.
This release supports the HP aCC compiler.
Three new icons have been added to allow the user to check out and check in source files within a source code repository. Each button is associated with a script to execute the appropriate command for retrieving or storing a source file. The default source repository is ClearCase but the scripts can be tailored to other systems, including RCS and SCCS.
This release contains a new utility, purela_show, which generates simplified reports from license and usage data maintained in a standard or aggregate PureLA database.
Purify has been tested with the following compilers:
See the "Restrictions and Known Issues" section for more details.
Purify supports these threads packages:
This change was made due to a problem in HP dld patches from June 1998 for HPUX 11.00 and August 1998 for HP-UX 10.20, which caused the instrumented application to hang after a call to shl_load(). One known instance of this problem occurs when you use getservbyname() because it loads network protocols using shl_load().
Linker and dld patches available after 3/99 don't have this problem. If the user has an earlier patch, an upgrade is strongly recommended.
Purify ships in 2 configurations, one supporting wide mode and the other supporting narrow. Both can be installed on the same file system, but the 64-bit version can only be used on 64-bit HP-UX 11.x systems.
If both install directories are in your path, Purify will auto-select the correct wide or narrow mode version, in most situations (see below for limitations). For example, you can install two versions of Purify:
purify-5.1-beta-H1-hpux (32-bit)
purify-5.1-beta-P1-hpux (64-bit)
(Beta and proto release with H<n>
in their name are 32-bit release.
P<n> signifies a 64-bit release.)
Or, for a final release:
purify-5.1-hpux (32-bit)
purify-5.1-hpux64 (64-bit)
If the two install directories are in your path, then running "purify" will automatically select the correct version, based on the type of program you are linking. The same is true for Quantify.
Even if only one install directory is on your path, auto-selection will occur if both versions are properly installed: Running pure_install on each version will prompt you for the location of the other product.
If you already have your licenses installed and do not choose to run pure_install, you can set up this connection between the two install directories by running the script "pure_link_32_64" in each install directory:
% cd purify-5.1-hpux
% pure_link_32_64
<answer the questions>
% cd purify-5.1-hpux64
% pure_link_32_64
Auto-selection only works between 32- and 64-bit Purify from 5.0 onwards.
In some situations, auto-selection does not have enough context to tell which version (32 or 64-bit) you need. This is true for the options:
-help
-printhomedir
-test-license
-version
In these cases, Purify defaults to the 32-bit version unless you explicitly specify what you want, using the new -ptr64 and -ptr32 options:
% purify -ptr64 -test-license
% quantify -ptr64 -printhomedir
% purify -ptr32 -version
% quantify -ptr32 -help
Failure to include -ptr<32|64> in these cases may yield the wrong information. For example, you may get the product home directory for the 32-bit product when you wanted the 64-bit product.
These options are NOT necessary during normal instrumentation and viewing operations:
% purify cc -g -o foo foo.o
% quantify -view my_app.qv
If you attempt to use the 64-bit Purify using -ptr64, or by having it on your PATH first, on a non 11.x systems, execution will fail. It only runs on HP-UX 11.x and later.
Because of a defect in auto-selection, auto-selection does not occur when using "-nolink". You must use -ptr64 or -ptr32 to ensure the correct version is used:
% quantify -ptr32(or -ptr64) -nolink ld mylib.a
You may also workaround this problem by include the following in your PUREOPTIONS environment variable:
-force-no-dynhash=no
The default setting of this option (to "yes") is used to workaround an HP bug in newer linkers.
b *purify_stop_here
When you stop at purify_stop_here in the debugger, the debugger's stack trace will not be correct: You won't see the stack trace for you program, only for Purify's runtime libraries.
The easiest way to get around this is to set a temporary breakpoint at the address which is in register r15 using this wdb command:
tb *$r15
Then, you can use the "continue" command to stop in your program at the instruction that caused the Purify error report.
Error: Child process exited with status = 1.
This is caused by an 'ar' failure in this locale. A workaround is to unsetenv LANG before instrumenting.
! Ignore the NumLock and ScrollLock keys on
! mouse buttons
Purify*ignoreModifierMask: Mod3|Mod2
This second workaround will take effect for a new Purify viewer after you restart your X-session or run a command like 'xrdb -merge $HOME/.Xdefaults'.
The site-wide URL for ClearQuest can be given during installation or set by manually editting the file
pure_clearquest_url
in your Purify home directory. A user can override the site-wide URL by setting the environment variable
PURE_CLEARQUEST_URL
This feature is partly implemented by a shell script, ("pure_invoke_clearquest" in your Purify home directory) to allow you to tailor its operation to your needs. If you wish, you may copy and customize this. script. As long as the directory containing the script appears in your search path before your Purify home directory, it will be used instead of the original script.
If you prefer to use Purify with ClearDDTS, you can do so by setting the X resource
Purify*ddtsCommandString
to 'xddts', if xddts is in your search path, or to the full path to your xddts executable. xddts is invoked by a shell script ("pure_invoke_ddts" in your Purify home directory). If you wish to customize it, please read the section on customizing "pure_invoke_clearquest" above.
If you already have a customized "pure_invoke_ddts" script in your search path, All you need to do is set your X resource as described above, and Purify will find your customized script automatically.
The following copyright applies to portions of this ClearQuest integration code:
Copyright 1996 Netscape Communications Corporation, all rights reserved. Created: Jamie Zawinski jwz@netscape.com, 24-Dec-94. Permission to use, copy, modify, distribute, and sell this software and its documentation for any purpose is hereby granted without fee, provided that the above copyright notice appear in all copies and that both that copyright notice and this permission notice appear in supporting documentation. No representations are made about the suitability of this software for any purpose. It is provided "as is" without express or implied warranty.
These compilers generate incorrect debugging information for some functions. Programs compiled using these compilers may not be debuggable after translation. In this case Purify may produce a message beginning "Failure relocating..." HP patch PHSS_4923 corrects this problem.
g++ 2.7.2 exceptions are not supported.
Some compilers load data from memory but ignore the data that has been read. Purify will signal a UMR if the loaded data is uninitialized. In some sense this is a false error report because the uninitialized data can not affect your program.
To avoid this kind of problem, you should run your application on a different X server than the Purify UI or Purify stderr output, or you should use the -log-file= or -view-file= options to specify a file to capture messages for inspection after your application is finished.
A convenient way to debug on two displays is to pre-start the Purify Viewer on one display ("slave"), and then start the application on the other display ("master"):
% purify -display slave:0 -view a.out.X &
% a.out.X -display master:0
The two commands must be executed on the same computer, but it could be the workstation associated with either display, or altogether another computer remote from both displays. The application will connect to the already started Purify Viewer, and messages will not conflict with the X display interactions of the application under test.
Wait...loading shared-library map tables.
xdb panic: Mapped addresses for dld
overlap text segment for dld
There is a simple workaround for this
problem and we've implemented it in the
shell script
prop system -on
alias `after_debug delete intercept signal SIGCHLD; \
prop system -off; \
breakpoint -in main -entry -exit; \
go
JIT debugging may fail to attach to your application if the executable resides on an NFS file system mounted without the "nointr" option. The HP Debugger reference manual says:
"If you get a Permission denied error message when you attach to a running process, it is likely that you are running either the debugger or the target process over an NFS link and that the relevant file system is mounted with the default intr option. You must mount the file system with the nointr option to resolve this problem. Use a command like the following to mount the file system containing the debugger:
mount -o nointr[,other_options] \ system:/opt/langtools /tools
Use a command like the following to mount the file system containing the target process:
mount -o nointr[,other_options] \ system:/test_area /test
It is probably easier to create an auxiliary mount for the file system than to unmount and remount it."
Purify does not support a type of relocation information known as "old style fixups". These were generated by HP-UX system software before release 3.0. If Purify detects old style fixups the message:
Object file has incompatible format
(may be older than HPUX 3.0)
is generated. We have seen this problem with HP's libsql.a and some of Oracle's Oracle6 libraries.
There is a simple workaround. Given a problem object module (or modules) the workaround is to have /bin/ld build a new object module. Suppose the old object modules are called `foo.o' and `bar.o'. Issuing the command:
% ld -r -o new_foo.o foo.o
% ld -r -o new_bar.o bar.o
or
% ld -r -o foo_and_bar.o foo.o bar.o
would generate a new object module where the old style fixups have been removed.
In the case of an archive file the following script will create a new archive given the full pathname of the original:
#!/bin/sh
# Remove old fixups from an archive.
# Supply original .a name as first argument.
cd /tmp
lib=new_`basename $1`
ar x $1
rm -f $lib
for member in `ar t $1` ; do
ld -r -o _$member $member
ar q $lib _$member
rm $member _$member
done
echo Created `pwd`/$lib