Huge thanks to our Platinum Members Endace and LiveAction,
and our Silver Member Veeam, for supporting the Wireshark Foundation and project.

Wireshark-dev: Re: [Wireshark-dev] Hmmm, don't understand why this failed

From: Guy Harris <guy@xxxxxxxxxxxx>
Date: Wed, 6 Mar 2019 19:33:03 -0800
On Mar 6, 2019, at 7:12 PM, Richard Sharpe <realrichardsharpe@xxxxxxxxx> wrote:

> I got this error with my latest petri-dish build:
> 
> 31: ======================================================================
> 31: FAIL: test_unit_fieldcount (suite_unittests.case_unittests)
> 31: fieldcount
> 31: ----------------------------------------------------------------------
> 31: Traceback (most recent call last):
> 31:   File "/home/wireshark/builders/ubuntu-x86-64-petri-dish/ubuntu-x86-64-petri-dish/build/test/fixtures.py",
> line 47, in wrapped
> 31:     test_fn(self, *fixtures)
> 31:   File "/home/wireshark/builders/ubuntu-x86-64-petri-dish/ubuntu-x86-64-petri-dish/build/test/suite_unittests.py",
> line 53, in test_unit_fieldcount
> 31:     self.assertRun((cmd_tshark, '-G', 'fieldcount'), env=test_env)
> 31:   File "/home/wireshark/builders/ubuntu-x86-64-petri-dish/ubuntu-x86-64-petri-dish/build/test/subprocesstest.py",
> line 289, in assertRun
> 31:     self.assertEqual(process.returncode, expected_return)
> 31: AssertionError: 1 != 0
> 31:
> 31: ----------------------------------------------------------------------
> 31: Ran 8 tests in 7.366s
> 31:
> 31: FAILED (failures=1)
> 
> Not sure what the error is. Can anyone help?

From a recent build:

	$ build/run/tshark -G fieldcount
	There are 215003 header fields registered, of which:
	        0 are deregistered
	        2533 are protocols
	        16078 have the same name as another field

	215000 fields were pre-allocated.
	* * Please increase PROTO_PRE_ALLOC_HF_FIELDS_MEM (in epan/proto.c)! * *

	The header field table consumes 1687 KiB of memory.
	The fields themselves consume 15117 KiB of memory.
	$ echo $?
	1

I boosted the #define in question to 220000 in

	https://code.wireshark.org/review/c/32345/

and it now works.