[PATCH] wireshark pidl related
mat at samba.org
Sun Oct 27 02:29:46 MDT 2013
On 10/27/2013 12:35 AM, Stefan (metze) Metzmacher wrote:
> Hi Matthieu,
>> I'm in a cycle to fix pidl for wireshark so that we can use recent idl
>> for exisiting dissector (atsvc, dnsserver, ...) but also new idl like
>> frstrans or drsuapi.
>> Here is a first set of patches for it:
> This patch is not needed, we have a helper function called
> ContainsPipe(), please use that one.
> As above.
>> From ac78b22ce8520cd9431e57ff2925317363e7260b Mon Sep 17 00:00:00 2001
>> From: Matthieu Patou <mat at matws.net>
>> Date: Tue, 1 Oct 2013 11:37:19 -0700
>> Subject: [PATCH 02/13] librpc: add idl for DFS-R (frsrtans) blobs
>> Signed-off-by: Matthieu Patou <mat at matws.net>
>> librpc/idl/frstransblobs.idl | 29 +++++++++++++++++++++++++++++
>> librpc/idl/wscript_build | 2 +-
>> librpc/wscript_build | 5 +++++
>> source4/librpc/wscript_build | 3 ++-
>> 4 files changed, 37 insertions(+), 2 deletions(-)
>> create mode 100644 librpc/idl/frstransblobs.idl
>> diff --git a/librpc/idl/frstransblobs.idl b/librpc/idl/frstransblobs.idl
>> new file mode 100644
>> index 0000000..b6ee30d
>> --- /dev/null
>> +++ b/librpc/idl/frstransblobs.idl
>> @@ -0,0 +1,29 @@
>> +#include "idl_types.h"
>> +//import "drsuapi.idl", "misc.idl", "samr.idl", "lsa.idl", "security.idl";
>> + uuid("12345778-1234-bcde-0001-000000000001"),
>> + version(0.0),
>> + pointer_default(unique),
>> + //helper("../librpc/ndr/ndr_drsblobs.h"),
>> + helpstring("DFS-R blobs")
> Please remove the unused stuff here.
Ok will do.
> The rest of the patches look ok from reading them,
> but I'd like to see the impact they have.
> For pidl changes it really got to maintain a branch
> where the generated files are commited and a reviewer is able to
> see the impact of each individual change, as there're often side effects
> which reveal bugs in inrelated code. Could you create such a branch
> maybe within the wireshark tree?
I have a wireshark tree, but it's after all my patches I haven't
submitted them for ease of review (and some of them need a bit of
rework, not that the generated output will change but I will use some
Trust me I know more than anybody else that when you try to do a fix for
the wireshark dissectors it's like playing "wack a mole": you make one
dissector generation better but you break four others.
I think this is the first time that with the whole patch set that we are
able to regenerate *all* the dissectors.
Honestly I think in this case it will be a bit tedious, some idl haven't
been recompiled for 5 years so the delta is pretty high !
You can have a look at my gitorious repo for wireshark:
But most of the regenerated patches are something like that
are you sure you really want to review changes like that ?
More information about the samba-technical