On 04/09/2019 11:05, Dave Martin wrote:
> On Fri, Aug 30, 2019 at 03:29:29PM +0100, Cristian Marussi wrote:
>> Hi
>>
>> On 13/08/2019 17:25, Dave Martin wrote:
>>> On Fri, Aug 02, 2019 at 06:02:55PM +0100, Cristian Marussi wrote:
>
> [...]
>
>>>> diff --git a/tools/testing/selftests/arm64/signal/testcases/fake_sigreturn_bad_magic.c b/tools/testing/selftests/arm64/signal/testcases/fake_sigreturn_bad_magic.c
>
> [...]
>
>>>> +static int fake_sigreturn_bad_magic_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 2*HDR_SZ space ... one for the KSFT_BAD_MAGIC
>>>> + * and the other for the usual terminator.
>>>> + */
>>>> + if (head && resv_sz - offset < HDR_SZ * 2) {
>>>
>>> Can we factor out this logic for finding space in the signal frame?
>>>
>>> We do pretty much the same thing in all the fake_sigreturn tests...
>>
>> Ok
>>>
>>>> + 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 */
>>>
>>> Do we actually time out? I don't see where we actually wait, so doesn't
>>> test_run() just fail immediately?
>>>
>>> The same applies to all the other fake_sigreturn tests too.
>>>
>> Right. It is probably a leftover.
>>
>> SIGALRM is used as an extreme measure to kill tests gone bad, but this
>> can happen only once the fake sigframe has been effectively placed on the stack
>> and sigreturned.
>
> OK, so this gets reported as a test failure because with no SIGSEGV,
> nothing ever sets td->pass?
Yes exactly. End result is based on value on td->pass, in case of abrupt
termination or timeout nobody sets td->pass ever.
>
> This is probably OK for now, though I wonder whether this should be
> reported as a skipped test instead.
>
> In case of doubt, reporting a failure is preferable anyway, since that
> will encourage people actually to investigate what went wrong.
>
As of now I never skip a test in fact...also tests for unsupported features
are built and run expecting a SIGILL, and reported as PASS in that case.
Cristian
> [...]
>
> Cheers
> ---Dave
>
