| 00:07.07 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 00:18.52 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 02:10.08 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 03:08.46 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 03:26.56 | *** join/#asterisk-dev acidfoo_ (~acidfoo@108.175.228.140) |
| 03:30.04 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 03:49.06 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 04:05.32 | *** join/#asterisk-dev klow (~textual@73.53.31.109) |
| 04:19.45 | *** join/#asterisk-dev cresl1n (~Adium@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 04:19.46 | *** mode/#asterisk-dev [+o cresl1n] by ChanServ |
| 05:00.09 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 05:13.14 | *** join/#asterisk-dev ccesario_ (~ccesario@email.logicotelecom.net.br) |
| 05:54.10 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 08:14.01 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 08:24.33 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 10:55.28 | *** join/#asterisk-dev ccesario_ (~ccesario@email.logicotelecom.net.br) |
| 11:45.51 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 11:46.49 | *** join/#asterisk-dev acidfoo_ (~acidfoo@108.175.228.140) |
| 13:06.46 | file | Fridayyyyyyy |
| 13:27.41 | *** join/#asterisk-dev leedm777 (leedm777@nat/digium/x-oxxbuphkxmwohnst) |
| 13:47.54 | leifmadsen | much friday |
| 13:47.59 | gtjoseph | Oh great. Woke up to 4 inches of snow. Expecting 5 more today. |
| 13:48.32 | gtjoseph | Good excuse to stay in and work on Asterisk. Oh wait... I do that anyway. |
| 13:52.26 | *** join/#asterisk-dev cresl1n (Adium@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 13:52.26 | *** mode/#asterisk-dev [+o cresl1n] by ChanServ |
| 13:55.43 | file | snow snow snow |
| 14:07.38 | gtjoseph | file: are you going to do the gerrit upgrade this weekend? |
| 14:07.45 | file | no |
| 14:08.02 | file | I have penciled it in for 9PM on March 28th |
| 14:08.05 | file | 9PM CST. |
| 14:08.13 | gtjoseph | gotcha |
| 14:08.24 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-wtojzlhwgyztqiyz) |
| 14:29.44 | gtjoseph | file: got a minute to look at ast_sip_push_task? |
| 14:29.49 | file | yes |
| 14:30.58 | gtjoseph | a few questions... What's the point of pulling a random serializer out of the pool if one's not specified. And... if we're pulling a random one out of the pool, why the check a few lines down? |
| 14:31.46 | file | read commit ada7346792452f021911063668997f79fdabc1f1 |
| 14:32.32 | file | that answers the first, and the second may have just been leftover from the original implementation |
| 14:33.42 | gtjoseph | hmmm. but don't the serializers use the same sip_threadpool anyway? |
| 14:33.52 | gtjoseph | looks |
| 14:33.57 | file | yes? |
| 14:35.42 | gtjoseph | so if they don't need to be run in sequence, why not go directly to sip_threadpool? |
| 14:35.52 | gtjoseph | i'm not following the significance of the serializer |
| 14:36.02 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-ndwwztyyjymyklch) |
| 14:36.21 | file | it ensures that any received traffic is not handled at the same time that the current thing is being handled |
| 14:36.52 | gtjoseph | mmmm. but if the serializer is chosen at random? |
| 14:37.40 | gtjoseph | the current thing may run in one serializer and the received traffic in another. |
| 14:37.45 | file | there exists functionality which allows you to query the serializer under which you are running and it gets associated |
| 14:37.58 | gtjoseph | Oh THAT! |
| 14:38.02 | file | so that anything coming back is put in that serializer and thus guaranteed to not execute at the same time |
| 14:38.47 | gtjoseph | Makes sense now. |
| 14:38.59 | gtjoseph | thanks |
| 14:41.16 | gtjoseph | wait..is that (ast_sip_thread_is_servant()? |
| 14:42.10 | file | that's a boolean, ast_threadpool_serializer_get_current gives you the name |
| 14:42.19 | file | well - the serializer |
| 14:42.24 | file | and then ast_taskprocessor_name the name |
| 14:43.02 | gtjoseph | hmm. i thought threadpool has a taskprocessor anyway. |
| 14:43.14 | gtjoseph | ok.. i'll investigate further. |
| 14:43.22 | file | what are you investigating? |
| 14:43.57 | gtjoseph | was thinking about the housecleaning threads and something to make them easier. |
| 14:44.10 | file | housecleaning threads? |
| 14:44.29 | file | the scheduler stuff and one-offs? |
| 14:44.32 | gtjoseph | registry expiration, etc. |
| 14:44.35 | file | ah |
| 14:45.00 | file | I would not advise messing in threadpool threading or serializer stuff... |
| 14:45.19 | gtjoseph | not messing in it, just wanting to understand it |
| 14:45.24 | file | ack |
| 14:45.58 | gtjoseph | haven't had much call to dive into it until now. |
| 14:46.49 | file | generally they are hidden and just do stuff and things |
| 14:47.16 | gtjoseph | i don't like black boxes. :) |
| 14:47.46 | file | if a black box does what I need reliably and well and lets me focus on other things, I'm all for 'em |
| 14:48.53 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-wtojzlhwgyztqiyz) |
| 14:50.49 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-ojajopjkqtoffyss) |
| 15:03.38 | *** part/#asterisk-dev lumasepa (~gestoip@193.145.124.30.local.ull.es) |
| 15:09.54 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-icgxtyaylsvovhpe) |
| 15:33.20 | *** join/#asterisk-dev acidfoo_ (~acidfoo@108.175.228.140) |
| 15:39.09 | gtjoseph | anyone know if the gcc cleanups (RAII_VAR, SCOPED_) are run in a deterministic order? Looks like LIFO but can't find any documentation. |
| 15:47.49 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-dvntyafsktvzljdb) |
| 16:53.50 | coreyfarrell | gtjoseph: only thing I'm seeing is that it's run when the var falls out of scope. given: 'char *v1;char *v2;' is it possible for v1 to be out of scope before v2? |
| 16:54.18 | gtjoseph | hmmm. |
| 16:55.24 | coreyfarrell | since the variables are actually stored on the stack is it even possible to go against LIFO? the stack is LIFO at the hardware level. |
| 16:55.24 | gtjoseph | if they're in the same stack frame, i think they go at the same time |
| 16:55.48 | gtjoseph | true but when does gcc run the cleanups? |
| 16:56.34 | gtjoseph | i.e before it pops the stack frame, does it run all the cleanups? |
| 16:58.34 | gtjoseph | i think what it does is push the cleanups as they're registered, then just before the stack frame pop, it pops all the cleanups |
| 16:58.45 | gtjoseph | which would make sense |
| 17:02.21 | coreyfarrell | gtjoseph: I do not think GCC would change the cleanup order in a way that could not be done with explicit scope brackets. so { char *v1; { char *v2; }; }; -- it's impossible to place brackets so the first declared variable falls out of scope first |
| 17:02.35 | gtjoseph | correct |
| 17:03.20 | gtjoseph | but out-of-scope may just mean when the stack frame they were allocated in is due to be popped. |
| 17:03.41 | gtjoseph | i don't think they come out of the stack individually |
| 17:04.08 | gtjoseph | that's why they have to be auto. |
| 17:04.49 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 17:07.51 | rmudgett | The RAII_VAR destructors are called when a variable goes out of scope. This includes RAII_VAR variables declared within a for loop. The destructor is called for each iteration of the for loop. |
| 17:08.25 | rmudgett | for (;;) { RAII_VAR(v1..); } |
| 17:09.07 | gtjoseph | yep. but if I do 3 RAII_VARs what order do the cleanups get run? |
| 17:10.44 | rmudgett | That may be implementation dependent. I have seen evidence that { v1; v2; v3; ... } executes destructors in v3, v2, v1 order |
| 17:10.49 | gtjoseph | or if I do RAII_VAR(some ao2 obj), then SCOPED_AO2LOCK(same ao2 obj), will the ao2_unlock run before the ao2_cleanup? |
| 17:11.55 | rmudgett | I would expect so. If it is that critical you can always add extra scope nesting. |
| 17:12.16 | rmudgett | That would force the expected order |
| 17:12.44 | coreyfarrell | rmudgett: wouldn't that violate scoping? I mean how can a compiler decide to do something that can't be expressed with explicit {} scoping? |
| 17:12.58 | gtjoseph | which is my approach since I couldn't find any docs to guarantee the unwind. |
| 17:13.17 | coreyfarrell | or to rephrase wouldn't it be a bug for a compiler to do that..? |
| 17:13.45 | gtjoseph | to do which? the unwind? |
| 17:13.54 | rmudgett | { v1; v2 ... } Both v1 and v2 fall out of scope at the same time |
| 17:14.06 | gtjoseph | exactly |
| 17:14.13 | rmudgett | {v1; ... {v2;...}} v2 falls out of scope before v1 |
| 17:23.00 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-179-40-172.cust.bezeqint.net) |
| 17:23.00 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 17:25.17 | *** join/#asterisk-dev pchero (~pchero@109.70.54.56) |
| 17:33.19 | file | ANYCAST IPS FOR ALL! |
| 17:34.17 | rmudgett | With anycast ips you can send a packet anywhere in the world. |
| 17:34.37 | rmudgett | Maybe even into the past |
| 17:35.29 | gtjoseph | i was trying to explain anycast to someone last week but they couldn't wrap their head around the same ip address being in more than 1 place. |
| 17:36.24 | rmudgett | Should rewatch Multiplicity |
| 17:36.41 | file | rmudgett, well if it crosses timezones... then it could go back in time |
| 17:37.06 | rmudgett | Na Just use a negative delay to send the packet |
| 17:49.10 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-dvntyafsktvzljdb) |
| 17:53.11 | *** join/#asterisk-dev fornax (~fornax@85.183.43.221) |
| 18:12.54 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 18:13.32 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-kynxrmogmgbpnfxp) |
| 18:53.31 | *** join/#asterisk-dev fornax (~fornax@85.183.43.221) |
| 18:56.52 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-drxwcdbjttlpdopg) |
| 19:30.58 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 19:58.42 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 21:26.27 | *** join/#asterisk-dev pchero (~pchero@109.70.54.56) |
| 21:43.13 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 22:33.49 | *** join/#asterisk-dev klow (~textual@c-73-53-31-109.hsd1.wa.comcast.net) |
| 22:45.16 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-drxwcdbjttlpdopg) |
| 23:19.39 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |