| Commit message (Collapse) | Author | Age | Files | Lines |
| ... | |
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19952 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19951 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
| |
Fixed so it does nothing when moving an element to its own index.
ticket: new
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19950 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
| |
autoconf.h, so that we don't need a redundant invocation of
config.status during the first build.
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19937 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
| |
remove the registration requirement for administrative installs.
ticket: new
component: windows
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19936 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
| |
behaves correctly and isn't acting on uninitialized variables.
ticket: 5704
tags: pullup
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19935 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
| |
so we get consistent errors. Expect the "keytab not found" error,
rather than the "principal doesn't exist" error.
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19931 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
| |
ticket: 4644
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19930 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
| |
AMD64 builds must load leashw64.dll not leashw32.dll
ticket:new
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19929 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
| |
initial patch. Memory written to by the IXDR macro calls had not been
accounted for. Thanks to Kevin Coffman, Will Fiveash, and Nico
Williams for discovering this bug and assisting with patch
development.
ticket: 5706
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19923 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add support for building 64-bit AMD64 MSI install packages
in addition to 32-bit i386 MSI install packages.
Differences between 32-bit MSI and 64-bit MSI include:
* no krb4 binaries and libraries
* no kclient binaries and libraries
* no krb524 binaries and libraries
* no leash32.exe
* new AMD64 UpgradeCode permits parallel installation with 32-bit MSI
* support for Visual Studio 2005 (VS8, CL1400) merge modules
Open Issues:
* 32-bit MSI installs kfwlogon.dll on AMD64 systems
* 32-bit and 64-bit NetIDMgr.exe as startup. Need to decide
which should be executed by default. Only one can run at
a time.
* Need to make sure that src/windows/build properly configures
the site-local.wxi file for Visual Studio 2005 and platform.
ticket: new
component: windows
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19922 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
| |
Now that krbcc64.lib exists we can build the gss.exe gss-api test
client.
ticket: new
component: windows
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19921 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
File locking was non-existant in this code and fccs chained list was
not used at all. This resulted in an assertion failure when closing the
cache. Code has been reorganized to parallel the code in krb5_fcc_resolve
for easier maintenence.
Commented out test in t_cc.c has been updated to actually test this code.
ticket:new
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19920 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
by only one mechanism. Revert RPC code to relying on this.
Build error-mapping code on a bidirectional map instead of a simple
array. When a status code is returned but has been seen returned from
a different mechanism already, generate a new number, starting at
100,000.
Use gssrpcint_printf for some more debugging code.
ticket: 5654
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19919 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19918 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19917 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19916 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
| |
ticket: 5707
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19915 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
| |
In kadm5_modify_policy_internal, check for nonexistence of policy
before doing anything with it, to avoid memory corruption.
ticket: new
target_version: 1.6.3
tags: pullup
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19914 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Make sure svcauth_gss_validate adequately checks oa->oa_length prior
to copying into rpcbuf.
ticket: new
target_version: 1.6.3
tags: pullup
component: krb5-libs
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19913 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
| |
ticket: 3334
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19911 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
| |
t_cc test case to pass and allows non-tickets to be stored (for now).
ticket: 5697
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19910 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
| |
The name of the Leash API DLL on Win64 is "leashw64.dll".
ticket: new
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19909 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19900 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19899 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
| |
names if they differ between the creds structure and the encoded
ticket and by looking up the server principal using the client's realm
if not found and server's realm was initially the referral (empty)
realm.
ticket: 5697
tags: pullup
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19898 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
NIM supports the ability of the user to specify an
explicit ccache name for use with an identity. If
this ccache is a FILE ccache, we need to be able to
store credentials into the ccache. krb5cred.dll
did not previously specify the KRB5_TC_OPENCLOSE flag
on the ccache when setting other flags such as
KRB5_TC_NOTICKET (which is used with MSLSA ccaches).
As a result, open/close mode was turned off, the
ccache file would be opened in read-only mode and
attempts to store credentials into the ccache would
fail. This is fixed by specifying KRB5_TC_OPENCLOSE
when setting the ccache flags.
When a CCAPI implementation is unavailable, we need
to automatically generate the FILE ccache name if
one has not already been specified. We default to
a file stored in the user's Local Settings\Temp
directory. The generated ccache is then added to
the file ccache watch list.
Finally, some users have complained about the
behavior of Microsoft Vista's UAC mode and how
it makes the CCAPI cache useless for storing
credentials that must be used in conjunction
with processes that do not have restricted
privileges since those processes run in a
separate logon session. For these users we
have added a "DefaultToFileCache" registry
value that can be specified to force the use
of FILE ccaches in preference to CCAPI ccaches
when there is no explicit ccache specified
for a given identity. Unlike CCAPI ccaches,
the FILE ccaches are accessible from both
restricted and unrestricted processes when
UAC is active.
ticket: new
component: windows
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19897 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
| |
Instead of testing for NOT the machine key, test
for is the user key.
ticket: new
component: windows/identity/kconfig/api.c
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19896 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
| |
Remove unused preprocessor symbols automatically added by Visual Studio's
dialog editor.
ticket: new
component: windows/identity/ui/resource.h
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19895 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
| |
New option to use krb5_sname_to_principal() for building principal
names, which is useful for testing referrals.
ticket: new
target_version: 1.6.3
tags: pullup
component: krb5-clients
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19894 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Split out the "standalone" test code from trval.c, so we produce
trval.o only once, instead of twice with different compilation flags.
One case generated the "trval" program directly, but produced and
deleted trval.o as an intermediate step in the compiler, which messes
things up if make thinks it's already built trval.o for another test
program.
Make the standalone test into t_trval.
Build problem reported by Will Fiveash of Sun, about ten minutes ago.
ticket: new
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19893 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
| |
The 64-bit CCAPI DLL name is krbcc64.dll not krbcc32.dll.
Conditionalize it on _WIN64.
ticket: new
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19892 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch permits Network Identity Manager to be built for 64-bit Windows.
In the process all compile time warnings have been taken care of.
For 64-bit Windows, we do not build the Kerberos v4 Credential Provider
and we will not attempt to load the krb524 library.
Note that when testing the 64-bit NIM, there is no CCAPI at the
moment so you must manually specify a FILE: ccache as part of the
identity's Kerberos v5 configuration if you want to use cache's
other than the MSLSA.
This patch also consolidates the computation of the default ccache
name into utility functions:
khm_krb5_get_identity_default_ccache
khm_krb5_get_identity_default_ccacheA
ticket: new
component: windows
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19891 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
| |
ticket: 2836
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19890 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
| |
multiple-inclusion protection of only sub-portions of k5-int.h,
presumably from a time when those portions were separate files. Since
the entire file is now protected against multiple inclusions, these
other macros aren't needed.
A few are left, mostly because they also appear in other files.
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19889 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19888 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
| |
to the function declarations enabled by KRB5_DEPRECATED. Definition
depends on having GCC version 3.2.3 or later. (Earlier versions may
have supported it, but that's what I found docs for. Windows compiler
support coming later, based on Jeff's suggestions.)
ticket: 2836
status: open
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19887 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
| |
ticket: 2836
status: open
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19886 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
| |
ticket: 5642
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19885 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19884 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
| |
ticket: 5670
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19883 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
| |
Oops.
ticket: 5670
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19882 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Microsoft defaults stack checking (/Gs) to on. This requires
that bufferoverflowU.lib be included in the link step. The
macro SCLIB in the build system specifies this library on
versions of Windows that require it. Include SCLIB on the
link line of the makefile.
ticket: new
component: windows
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19881 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
| |
choice_pa_pk_as_rep_draft9_UNKNOWN to cleanup warning... Value of enum
same in both cases...
ticket: 5617
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19880 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
| |
to be output.
ticket: 5657
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19879 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19878 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19877 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19876 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
|
|
|
|
|
| |
not used outside the krb5 library. (In some cases, the only use is in
our own test programs, which means two things: The test programs
should probably be changed to use the accessor interface, and some of
the newer encoder/decoder functions aren't being unit-tested.)
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19875 dc483132-0cff-0310-8789-dd5450dbe970
|
| |
|
|
| |
git-svn-id: svn://anonsvn.mit.edu/krb5/trunk@19874 dc483132-0cff-0310-8789-dd5450dbe970
|