tests/tcg/multiarch: Give the 'memory' TCG test a larger timeout

[PATCH] tests/tcg/multiarch: Give the 'memory' TCG test a larger timeout

Posted by Peter Maydell 9 months ago

The 'memory' TCG test times out intermittently on our cross-i686-tci
CI job. We expect this to be a slow config (it's using TCI), so it's
possible that when the CI runner is heavily loaded it tips past
the timeout.

Double the timeout for tests.  If this doesn't resolve the
intermittents we can assume we're looking at some kind of
TCI-specific hang.

Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2079
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
I could not figure out a way to get the timeout to apply to the
'memory' test specifically (including all its variants like
running with plugins or the gdbserver or record-replay).
---
 tests/tcg/Makefile.target | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tests/tcg/Makefile.target b/tests/tcg/Makefile.target
index 8cf65f68dd8..af1a18cee1b 100644
--- a/tests/tcg/Makefile.target
+++ b/tests/tcg/Makefile.target
@@ -98,7 +98,8 @@ QEMU_OPTS=
 #   15s    original default
 #   60s    with --enable-debug
 #   90s    with --enable-tcg-interpreter
-TIMEOUT=90
+# The 'memory' test in particular is very slow under TCI.
+TIMEOUT=180
 
 ifeq ($(filter %-softmmu, $(TARGET)),)
 # The order we include is important. We include multiarch first and
-- 
2.34.1

Re: [PATCH] tests/tcg/multiarch: Give the 'memory' TCG test a larger timeout

Posted by Peter Maydell 8 months, 3 weeks ago

Ping for review?

thanks
-- PMM

On Tue, 27 Feb 2024 at 14:23, Peter Maydell <peter.maydell@linaro.org> wrote:
>
> The 'memory' TCG test times out intermittently on our cross-i686-tci
> CI job. We expect this to be a slow config (it's using TCI), so it's
> possible that when the CI runner is heavily loaded it tips past
> the timeout.
>
> Double the timeout for tests.  If this doesn't resolve the
> intermittents we can assume we're looking at some kind of
> TCI-specific hang.
>
> Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2079
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
> I could not figure out a way to get the timeout to apply to the
> 'memory' test specifically (including all its variants like
> running with plugins or the gdbserver or record-replay).
> ---
>  tests/tcg/Makefile.target | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/tests/tcg/Makefile.target b/tests/tcg/Makefile.target
> index 8cf65f68dd8..af1a18cee1b 100644
> --- a/tests/tcg/Makefile.target
> +++ b/tests/tcg/Makefile.target
> @@ -98,7 +98,8 @@ QEMU_OPTS=
>  #   15s    original default
>  #   60s    with --enable-debug
>  #   90s    with --enable-tcg-interpreter
> -TIMEOUT=90
> +# The 'memory' test in particular is very slow under TCI.
> +TIMEOUT=180
>
>  ifeq ($(filter %-softmmu, $(TARGET)),)
>  # The order we include is important. We include multiarch first and
> --

Re: [PATCH] tests/tcg/multiarch: Give the 'memory' TCG test a larger timeout

Posted by Alex Bennée 8 months, 3 weeks ago

Peter Maydell <peter.maydell@linaro.org> writes:

> Ping for review?

Ahh we crossed streams because I merged:

  cdb5bfc9f34 (tests/tcg: bump TCG test timeout to 120s)

Do we need even more time?

>
> thanks
> -- PMM
>
> On Tue, 27 Feb 2024 at 14:23, Peter Maydell <peter.maydell@linaro.org> wrote:
>>
>> The 'memory' TCG test times out intermittently on our cross-i686-tci
>> CI job. We expect this to be a slow config (it's using TCI), so it's
>> possible that when the CI runner is heavily loaded it tips past
>> the timeout.
>>
>> Double the timeout for tests.  If this doesn't resolve the
>> intermittents we can assume we're looking at some kind of
>> TCI-specific hang.
>>
>> Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2079
>> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
>> ---
>> I could not figure out a way to get the timeout to apply to the
>> 'memory' test specifically (including all its variants like
>> running with plugins or the gdbserver or record-replay).
>> ---
>>  tests/tcg/Makefile.target | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/tests/tcg/Makefile.target b/tests/tcg/Makefile.target
>> index 8cf65f68dd8..af1a18cee1b 100644
>> --- a/tests/tcg/Makefile.target
>> +++ b/tests/tcg/Makefile.target
>> @@ -98,7 +98,8 @@ QEMU_OPTS=
>>  #   15s    original default
>>  #   60s    with --enable-debug
>>  #   90s    with --enable-tcg-interpreter
>> -TIMEOUT=90
>> +# The 'memory' test in particular is very slow under TCI.
>> +TIMEOUT=180
>>
>>  ifeq ($(filter %-softmmu, $(TARGET)),)
>>  # The order we include is important. We include multiarch first and
>> --

-- 
Alex Bennée
Virtualisation Tech Lead @ Linaro

Re: [PATCH] tests/tcg/multiarch: Give the 'memory' TCG test a larger timeout

Posted by Peter Maydell 8 months, 3 weeks ago

On Mon, 4 Mar 2024 at 18:26, Alex Bennée <alex.bennee@linaro.org> wrote:
>
> Peter Maydell <peter.maydell@linaro.org> writes:
>
> > Ping for review?
>
> Ahh we crossed streams because I merged:
>
>   cdb5bfc9f34 (tests/tcg: bump TCG test timeout to 120s)
>
> Do we need even more time?

Ah, I missed that -- I'll let you know if I see this failure again.

-- PMM