| 06:28.21 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01:e0a:d:1111:b749:768d:ed77:7c9a) |
| 12:20.34 | *** join/#asterisk-dev jkroon (~jkroon@2c0f:f720:fe0c:1:f4e6:39d3:b9a7:8064) |
| 12:31.47 | *** join/#asterisk-dev linuxlad (~crosiak@74.115.41.9) |
| 13:19.24 | seanbright | is there a distribution out there where you compile everything yourself? |
| 13:19.31 | seanbright | is that what slackware or arch is? |
| 13:19.45 | seanbright | is slackware still a thing or am i dating myself? |
| 13:20.31 | gtjoseph | iirc, yes. everything gets downloaded as source and compiles at package install time. |
| 13:21.05 | seanbright | awesome |
| 13:22.31 | jsmith | seanbright: What, not Gentoo? |
| 13:22.37 | jsmith | has no love for Gentoo, just for the record. |
| 13:24.06 | seanbright | i know nothing about linux if i can't "yum install" or "apt install" |
| 13:24.54 | seanbright | my goal is to build software that can take better advantage of my processor |
| 13:25.04 | seanbright | rather than a 'generic' x86_64 one |
| 13:27.18 | seanbright | more software needs runtime cpu feature detection. but it's probably easier to just re-compile stuff. |
| 13:43.48 | jsmith | seanbright: Good luck with that... especially if you care about reproducability |
| 13:44.16 | seanbright | i don't care about that |
| 13:44.19 | seanbright | i care about speeeeeeeeed |
| 13:44.26 | seanbright | POWERRRRR |
| 13:44.37 | file | overclocks seanbright |
| 13:45.02 | seanbright | did some simple sampling with perf and found that srtp was doing some AES/SHA1 stuff |
| 13:45.14 | seanbright | well, my processor has instructions for those operations, but they weren't getting used |
| 13:45.19 | jsmith | seanbright: If your time is worth anything, it's cheaper to buy faster hardware. |
| 13:45.30 | seanbright | the hardware _is_ fast |
| 13:45.32 | jsmith | seanbright: Or rebuild RPM/deb packages *just* for the things you're really concerned about. |
| 13:45.36 | seanbright | sure |
| 13:45.53 | seanbright | i think rebuilding the srtp package is going to be my first step |
| 13:46.06 | jsmith | Seems like a very reasonable first step. |
| 13:46.12 | seanbright | well it was your idea |
| 13:46.24 | *** join/#asterisk-dev sduthil (~sduthil@bras-base-mtrlpq4806w-grc-11-69-157-142-95.dsl.bell.ca) |
| 13:47.44 | file | ideas can be a dangerous thing |
| 13:54.35 | jsmith | seanbright: Out of curiosity, which version of libsrtp are you using, and against which version of OpenSSL? |
| 13:55.03 | seanbright | jsmith: it's not linked against openssl, which is the problem |
| 13:55.03 | jsmith | seanbright: If you're using modern versions of both, you should get support for things like AES-GCM and AES-NI baked in. Or at least, that's what the interwebs lead me to believe. |
| 13:55.11 | jsmith | seanbright: Ah, there's your problem :-) |
| 13:55.18 | seanbright | the ubuntu version does not link against it |
| 13:55.26 | jsmith | Gotcha |
| 13:55.48 | seanbright | debian sid version just started linking against NSS, but i'm not going to be able to get that in focal |
| 14:04.46 | *** join/#asterisk-dev bford (uid283514@gateway/web/irccloud.com/x-bthmdktduejfbjxv) |
| 14:04.46 | *** mode/#asterisk-dev [+o bford] by ChanServ |
| 14:19.57 | *** join/#asterisk-dev kharwell (uid358942@gateway/web/irccloud.com/x-hfhlzertpmrlgieh) |
| 14:19.57 | *** mode/#asterisk-dev [+o kharwell] by ChanServ |
| 15:20.45 | *** join/#asterisk-dev luck_1122 (d9915757@217.145.87.87) |
| 15:21.17 | luck_1122 | hi guys |
| 15:23.06 | luck_1122 | where can be the problem features map: ** and *2 to work with Dial(SIP/1112,,tThH) but not in Queue(112233,HhkKtT,,,120,,,,,) |
| 15:23.50 | luck_1122 | i can see the DTMF in the console but the Queue still running and play the music on hold |
| 15:57.07 | seanbright | jsmith: i hate the "throw faster hardware at it" solution. computers are fast. |
| 15:57.49 | seanbright | but i don't disagree that my time is worth more than the expense of faster hardware |
| 16:47.35 | *** join/#asterisk-dev CELYA_ (~Thunderbi@2a01cb0583d31d00f4d7ddd1bc0987bc.ipv6.abo.wanadoo.fr) |
| 17:04.17 | *** join/#asterisk-dev CELYA_1 (~Thunderbi@lfbn-nan-1-2163-147.w90-105.abo.wanadoo.fr) |
| 19:23.59 | *** join/#asterisk-dev coreyfarrell (~coreyfarr@rrcs-66-57-14-122.midsouth.biz.rr.com) |
| 19:23.59 | *** mode/#asterisk-dev [+o coreyfarrell] by ChanServ |
| 19:49.55 | coreyfarrell | does the second argument given to stasis_message_create ever need to be allocated with a lock or is that object expected to be immutable? |
| 19:52.55 | file | that's up to the user of the message |
| 19:54.54 | coreyfarrell | file: up to the user (not the producer)... so removing an unused lock from a payload structure declared in a header could be a breaking change? |
| 19:55.10 | file | I more meant both |
| 19:55.29 | file | it's dependent on the contract between producer and consumer as to the usage of it |
| 19:55.37 | file | if that contract does not need a lock, then it's fine |
| 20:00.05 | coreyfarrell | what if it's not specified? take ast_json_payload_create for example, it creates an instance of a public structure but doesn't say anything about locking or not, just says it's an ao2 object. |
| 20:00.49 | file | then case by case probably? most likely immutable |
| 22:01.08 | *** join/#asterisk-dev pchero (~pchero@2a02:a210:2241:6480:1884:67c6:3748:c281) |