Re: [lng-odp] [API-NEXT PATCHv9 00/15] using _ishm as north API mem allocator

v9 merged, discussed issues we cover in separate patches. Maxim. On 11/25/16 15:03, Christophe Milard wrote: Hi Maxim, I thought a sighup signal would be sent to children when a process is killed, but I just tried, and it it not true. Only true for shells. I'll look for a solution... Good

Re: [lng-odp] [API-NEXT PATCHv9 00/15] using _ishm as north API mem allocator

Hi Maxim, I thought a sighup signal would be sent to children when a process is killed, but I just tried, and it it not true. Only true for shells. I'll look for a solution... Good catch, but not related to this patch series. Christophe On 25 November 2016 at 10:35, Maxim Uvarov

Re: [lng-odp] [API-NEXT PATCHv9 00/15] using _ishm as north API mem allocator

On 24 November 2016 at 21:07, Maxim Uvarov wrote: > Christophe, > > with this patch series and without it I see odp_pktio_perf hangs in > background. > And it looks like accept() never pass. Interesting is that Cunit reported > this test as > passed and actually it's in

Re: [lng-odp] [API-NEXT PATCHv9 00/15] using _ishm as north API mem allocator

Christophe, with this patch series and without it I see odp_pktio_perf hangs in background. And it looks like accept() never pass. Interesting is that Cunit reported this test as passed and actually it's in background. Other apps looks like ok. 81../sysdeps/unix/syscall-template.S: No

[lng-odp] [API-NEXT PATCHv9 00/15] using _ishm as north API mem allocator

since v8: - commit msg typo fix (Maxim) - use snprintf() instead of sprintf() (Maxim) - better handling of error cases (Maxim) since v7: - Fix not to use single-va memory when the requirement alignment exactely matches the page size (Maxim) - patch to clean-up all