Re: [PATCH v3 07/11] kselftest: arm64: fake_sigreturn_bad_size_for_magic0

[Cristian Marussi <cristian.marussi@arm.com>]

On 13/08/2019 17:25, Dave Martin wrote:
> On Fri, Aug 02, 2019 at 06:02:56PM +0100, Cristian Marussi wrote:
>> Added a simple fake_sigreturn testcase which builds a ucontext_t
> 
> Add

Ok
> 
>> with a badly sized magic0 header and place it onto the stack.
> 
> I usually call a record with magic number 0 a "terminator record".
> 

Ok
>> Expects a SIGSEGV on test PASS.
>>
>> Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
>> ---
>>  .../arm64/signal/testcases/.gitignore         |  1 +
>>  .../fake_sigreturn_bad_size_for_magic0.c      | 57 +++++++++++++++++++
>>  2 files changed, 58 insertions(+)
>>  create mode 100644 tools/testing/selftests/arm64/signal/testcases/fake_sigreturn_bad_size_for_magic0.c
>>
>> diff --git a/tools/testing/selftests/arm64/signal/testcases/.gitignore b/tools/testing/selftests/arm64/signal/testcases/.gitignore
>> index 0ea6fdc3765c..cf2a73599818 100644
>> --- a/tools/testing/selftests/arm64/signal/testcases/.gitignore
>> +++ b/tools/testing/selftests/arm64/signal/testcases/.gitignore
>> @@ -5,3 +5,4 @@ mangle_pstate_invalid_mode_el2
>>  mangle_pstate_invalid_mode_el3
>>  mangle_pstate_ssbs_regs
>>  fake_sigreturn_bad_magic
>> +fake_sigreturn_bad_size_for_magic0
>> diff --git a/tools/testing/selftests/arm64/signal/testcases/fake_sigreturn_bad_size_for_magic0.c b/tools/testing/selftests/arm64/signal/testcases/fake_sigreturn_bad_size_for_magic0.c
>> new file mode 100644
>> index 000000000000..2f53c4740c85
>> --- /dev/null
>> +++ b/tools/testing/selftests/arm64/signal/testcases/fake_sigreturn_bad_size_for_magic0.c
>> @@ -0,0 +1,57 @@
>> +/* SPDX-License-Identifier: GPL-2.0 */
>> +/* Copyright (C) 2019 ARM Limited */
>> +
> 
> #include <signal.h> ?
> 

Ok
>> +#include <stdio.h>
>> +#include <ucontext.h>
>> +
>> +#include "test_signals_utils.h"
>> +#include "testcases.h"
>> +
>> +struct fake_sigframe sf;
>> +
>> +#define MIN_SZ_ALIGN	16
>> +
>> +static int fake_sigreturn_bad_size_for_magic0_run(struct tdescr *td,
>> +						  siginfo_t *si, ucontext_t *uc)
>> +{
>> +	size_t resv_sz, offset;
>> +	struct _aarch64_ctx *shead = GET_SF_RESV_HEAD(sf), *head;
>> +
>> +	/* just to fill the ucontext_t with something real */
>> +	if (!get_current_context(td, &sf.uc))
>> +		return 1;
>> +
>> +	resv_sz = GET_SF_RESV_SIZE(sf);
>> +	/*
>> +	 * find the terminator, preserving existing headers
>> +	 * and verify amount of spare room in __reserved area.
>> +	 */
>> +	head = get_terminator(shead, resv_sz, &offset);
>> +	/*
>> +	 * try stripping extra_context header when low on space:
>> +	 * we need at least HDR_SZ + 16 space for the bad sized terminator.
>> +	 */
>> +	if (head && resv_sz - offset < HDR_SZ + MIN_SZ_ALIGN) {
>> +		fprintf(stderr, "Low on space:%zd. Discarding extra_context.\n",
>> +			resv_sz - offset);
>> +		head = get_header(shead, EXTRA_MAGIC, resv_sz, &offset);
>> +	}
>> +	/* just give up and timeout if still not enough space */
>> +	if (head && resv_sz - offset >= HDR_SZ + MIN_SZ_ALIGN) {
>> +		head->magic = 0;
>> +		head->size = MIN_SZ_ALIGN;
> 
> This is different from the amount of space we tested for
> (HDR_SZ + MIN_SZ_ALIGN) earlier.
> 
> I'm not sure it matters which we use, but we should be consistent.
> 
> I suggest sticking with HDR_SZ, unless there's something I've missed.
> 
I'll stick to HDR_SZ in this case and use the new helper from 06/11
(get_starting_head)

> [...]
> 
> Cheers
> ---Dave
> 

Cheers

Cristian