Am 06.02.2018 um 19:18 schrieb Jonathan Wakely: > On 6 February 2018 at 18:03, Rainer Emrich wrote: > >> At least 20 of the acats tests catch all memory until the host memory is >> exhausted. >> Same holds for the two libstdc++ tests >> 23_containers/unordered_set/requirements/exception/basic.cc >> and >> 23_containers/unordered_set/requirements/exception/propagation_consistent.cc > > Are these new failures? What changed? Where does it get stuck? >
Some of my observations, I don't know whats expected:
In the function populate p(container_control)
"make_container_n<container_type> made(n);" is called with n=82.
This allocates nearly 16GByte on memory. Is this expected?
Before the call to run_steps_to_limit(container_control, f) the
container_control structure has:
_M_bucket_count = 2019773507
_M_element_count = 82
_M_rehash_policy = {static _S_growth_factor = 2, _M_max_load_factor = 1,
_M_next_resize = 2019773507}
To answer my own question, that's not expected. So at this state it's
already wrong.
signature.asc
Description: OpenPGP digital signature
