| 00:23.13 | *** join/#asterisk-dev infobot (~infobot@rikers.org) |
| 00:23.13 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Check out our blog! blogs.asterisk.org -=- Follow on Twitter at @AsteriskDev |
| 01:07.35 | *** join/#asterisk-dev joshelson (~joshelson@70.102.238.132) |
| 01:59.30 | *** join/#asterisk-dev joshelson (~joshelson@70.102.238.132) |
| 02:07.53 | *** join/#asterisk-dev tm1000_ (sid6728@gateway/web/irccloud.com/x-qqqkfiwjledeexex) |
| 02:12.50 | *** join/#asterisk-dev zopsi_ (~zopsi@dir.ac) |
| 02:18.21 | *** join/#asterisk-dev joshelson (~joshelson@70.102.238.132) |
| 02:41.36 | *** join/#asterisk-dev joshelson (~joshelson@73.98.162.212) |
| 02:56.56 | *** join/#asterisk-dev Bordr (~Bordr@c-24-9-55-138.hsd1.co.comcast.net) |
| 04:26.44 | *** join/#asterisk-dev joshelson (~joshelson@70.102.238.132) |
| 06:53.31 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 06:54.58 | *** join/#asterisk-dev tsearle (~tsearle@37.19.10.33) |
| 06:54.58 | *** mode/#asterisk-dev [+o tsearle] by ChanServ |
| 06:57.33 | *** join/#asterisk-dev CELYA_ (~Thunderbi@LStLambert-657-1-90-73.w80-13.abo.wanadoo.fr) |
| 07:51.01 | *** join/#asterisk-dev oej (~oej@h-205-12.A165.corp.bahnhof.se) |
| 07:51.01 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 08:28.02 | *** join/#asterisk-dev oej (~oej@h-205-12.A165.corp.bahnhof.se) |
| 08:28.02 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 08:37.52 | *** join/#asterisk-dev tsearle (~tsearle@37.19.10.33) |
| 08:37.53 | *** mode/#asterisk-dev [+o tsearle] by ChanServ |
| 08:47.02 | *** join/#asterisk-dev tsearle (~tsearle@37.19.10.33) |
| 08:47.02 | *** mode/#asterisk-dev [+o tsearle] by ChanServ |
| 12:13.02 | *** join/#asterisk-dev snuff-work (~snuff-wor@210.9.148.102) |
| 12:13.03 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 12:37.01 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 12:37.01 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 12:57.23 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |
| 13:05.30 | *** join/#asterisk-dev libardi (~libardi@179.159.11.133) |
| 13:08.33 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |
| 13:25.43 | *** join/#asterisk-dev pc-m (~pcm@modemcable094.94-70-69.static.videotron.ca) |
| 13:33.50 | *** join/#asterisk-dev newtonr (newtonr@nat/digium/x-cpmjsopjnodqhwpp) |
| 13:33.50 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 13:36.30 | *** join/#asterisk-dev leedm777 (textual@nat/digium/x-gueouhuivfrccpbe) |
| 13:36.30 | *** mode/#asterisk-dev [+o leedm777] by ChanServ |
| 13:46.54 | *** join/#asterisk-dev CELYA_ (~Thunderbi@37.157.229.98) |
| 13:48.49 | *** join/#asterisk-dev cresl1n (Adium@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 13:48.49 | *** mode/#asterisk-dev [+o cresl1n] by ChanServ |
| 13:51.55 | *** join/#asterisk-dev CELYA_1 (~Thunderbi@37.157.229.127) |
| 13:53.38 | *** join/#asterisk-dev oej (~oej@h-205-12.A165.corp.bahnhof.se) |
| 13:53.38 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 14:00.21 | *** join/#asterisk-dev CELYA_ (~Thunderbi@37.157.229.98) |
| 14:16.42 | *** join/#asterisk-dev bford (d8cff501@gateway/web/freenode/ip.216.207.245.1) |
| 14:16.42 | *** mode/#asterisk-dev [+o bford] by ChanServ |
| 14:16.55 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-tsxuqgppwtartfsh) |
| 14:16.55 | *** mode/#asterisk-dev [+o kharwell] by ChanServ |
| 14:19.16 | *** join/#asterisk-dev oej (~oej@h-205-12.A165.corp.bahnhof.se) |
| 14:19.33 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 14:21.45 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-sllfhigmjqnpqnro) |
| 14:21.45 | *** mode/#asterisk-dev [+o rmudgett] by ChanServ |
| 14:35.17 | *** join/#asterisk-dev CELYA_ (~Thunderbi@LStLambert-657-1-90-73.w80-13.abo.wanadoo.fr) |
| 14:59.00 | seanbright | i'm still not in love with 5749 |
| 14:59.47 | *** join/#asterisk-dev leedm777 (textual@nat/digium/x-oxyvhbgtgeeqznql) |
| 14:59.47 | *** mode/#asterisk-dev [+o leedm777] by ChanServ |
| 15:00.11 | seanbright | for example, AST_EXTENSION_BUSY is still considered "closed" |
| 15:43.35 | coreyfarrell | anyone else having trouble with asterisk.org signin? |
| 15:44.53 | coreyfarrell | hmm it worked on the second attempt, very slow though. |
| 15:45.26 | gtjoseph | yeah, it's intermittent and we've reported it |
| 15:54.23 | coreyfarrell | cool thanks |
| 15:57.28 | *** join/#asterisk-dev leedm777 (textual@nat/digium/x-gwuwckjlxzrifgpd) |
| 15:57.28 | *** mode/#asterisk-dev [+o leedm777] by ChanServ |
| 16:05.23 | *** join/#asterisk-dev infobot (~infobot@rikers.org) |
| 16:05.23 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Check out our blog! blogs.asterisk.org -=- Follow on Twitter at @AsteriskDev |
| 16:05.26 | *** join/#asterisk-dev pc-m (~pcm@modemcable094.94-70-69.static.videotron.ca) |
| 16:11.30 | *** join/#asterisk-dev infobot (~infobot@rikers.org) |
| 16:11.30 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Check out our blog! blogs.asterisk.org -=- Follow on Twitter at @AsteriskDev |
| 16:35.04 | *** join/#asterisk-dev infobot (~infobot@rikers.org) |
| 16:35.04 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Check out our blog! blogs.asterisk.org -=- Follow on Twitter at @AsteriskDev |
| 18:26.10 | *** join/#asterisk-dev infobot (~infobot@rikers.org) |
| 18:26.10 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Check out our blog! blogs.asterisk.org -=- Follow on Twitter at @AsteriskDev |
| 18:28.45 | *** join/#asterisk-dev luckman212 (~luckman21@unaffiliated/luckman212) |
| 18:39.39 | seanbright | file: can you explain put_unaligned_uint32() to me like i am a child? |
| 18:40.11 | seanbright | or maybe just an adolescent |
| 18:40.23 | file | my brain is mush |
| 18:40.26 | file | rmudgett: you explain! |
| 18:41.32 | file | voluntelling didn't seem to work |
| 18:41.35 | rmudgett | Processors like to write 32 bit quantities on 32 bit boundaries. |
| 18:42.26 | rmudgett | When where you want to write the 32 bit quantity doesn't align with that boundary it is unaligned. :) |
| 18:42.57 | seanbright | thanks, that clears things up |
| 18:42.57 | rmudgett | In that case you have to break up the write into two or more writes. |
| 18:43.16 | gtjoseph | and reads |
| 18:43.29 | seanbright | just so that i am clear... if i have a unsigned char * to an arbitrary memory location that is not aligned on a 4 byte boundary |
| 18:43.36 | rmudgett | Yes those two but that wasn't what was asked. :D |
| 18:44.23 | seanbright | and i do: *((unsigned int *) p) = 128; |
| 18:44.23 | seanbright | it won't use the 4 byte pointed to by p? |
| 18:44.28 | seanbright | bytes* |
| 18:44.54 | gtjoseph | casting pointer to int is VERY dangerous |
| 18:45.03 | rmudgett | On x86 based processors the processor will break the write into several writes. |
| 18:45.30 | rmudgett | On processors like the Motorola 68000 you will get a segfault |
| 18:46.01 | rmudgett | Or that should be a bus error. |
| 18:46.16 | seanbright | so how does casting the pointer to a struct pointer and then dereferencing work around that? |
| 18:46.49 | rmudgett | What code are you looking at? |
| 18:47.08 | seanbright | https://github.com/asterisk/asterisk/blob/95cf4f8d3128d3c972126efd5b7e4e7b55486fa2/include/asterisk/unaligned.h#L58 |
| 18:47.39 | seanbright | ahh, that's the GCC extension |
| 18:48.09 | seanbright | line 115 has the split write you mentioned |
| 18:49.19 | gtjoseph | not really, that's just shifting the bits |
| 18:49.46 | rmudgett | The key with the GCC code you point to is the packed attribute. |
| 18:49.52 | seanbright | it's writing the data 1 byte at a time |
| 18:50.00 | seanbright | instead of treating it as an int |
| 18:50.06 | seanbright | how is that not a "split write?" |
| 18:51.01 | gtjoseph | i was thinking what instructions the compiler generates, not what the app does. |
| 18:51.22 | gtjoseph | so technically, yes it's a split write |
| 18:51.29 | seanbright | super |
| 18:51.43 | gtjoseph | but it's better to let the compiler do its think with the _packeed attribute |
| 18:51.57 | rmudgett | The sparc processor does not allow misaligned writes similar to the 68000 processors. |
| 18:52.19 | seanbright | so what does the packed attribute do that a pointer-to-int would not? |
| 18:52.27 | seanbright | changes which instructions the compiler generates? |
| 18:52.43 | gtjoseph | it takes care of doing the split write yourself |
| 18:52.55 | gtjoseph | you set it on the structure |
| 18:53.08 | gtjoseph | and the compiler does the work |
| 18:53.53 | seanbright | i guess i just don't understand |
| 18:53.57 | rmudgett | I wonder if the compiler only allow using that attribute on structs which would be why a struct is involved. |
| 18:54.12 | seanbright | i don't see anything in the documentation for 'packed' that has anything to do with alignment |
| 18:54.50 | gtjoseph | iirc you CAN use it elsewhere but where would the savings be? |
| 18:54.51 | rmudgett | Packed structs involve misaligned members because padding is not allowed. |
| 18:55.30 | seanbright | ok... but in what circumstances would casting to a pointer-to-int _not_ work? |
| 18:55.43 | seanbright | on x86 |
| 18:56.16 | gtjoseph | i guess it would work fine but the question remains...why? and in what context? |
| 18:56.37 | seanbright | who asked those two questions? |
| 18:56.48 | gtjoseph | I just did :) |
| 18:57.11 | seanbright | maybe i need to rephrase... |
| 18:57.24 | seanbright | well, maybe i don't care anymore |
| 18:57.26 | file | offers seanbright some water |
| 18:57.42 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-179-40-170.cust.bezeqint.net) |
| 18:57.43 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 18:57.52 | seanbright | i'm trying to understand why a pointer-to-struct is any better than a pointer-to-int |
| 18:58.03 | seanbright | since they will both be unaligned |
| 18:58.30 | gtjoseph | there's no context in which to answer that question. better than what? |
| 18:58.42 | seanbright | huh? |
| 18:58.48 | seanbright | re-reads the scrollback |
| 18:58.55 | seanbright | gtjoseph: let me get you caught up |
| 18:59.13 | rmudgett | There are only two modules using put_unaligned_uint32(): astmm and chan_iax2 |
| 18:59.23 | seanbright | gtjoseph: https://pastebin.com/raw/MxnAsc6h |
| 18:59.28 | gtjoseph | i got it. |
| 18:59.33 | seanbright | and res_rtp_asterisk |
| 19:00.00 | gtjoseph | ok, so let's look at res_rtp_asterisk |
| 19:00.05 | gtjoseph | where's it used? |
| 19:00.26 | seanbright | huh? |
| 19:00.33 | seanbright | ok... i'm slowly backing out of the room now |
| 19:00.55 | seanbright | will PM rmudgett if i have additional questions |
| 19:01.09 | seanbright | this leading-a-man-to-water schtick is obnoxious |
| 19:01.09 | gtjoseph | the point i twas going to make was "unaligned.h" may not be good |
| 19:02.10 | seanbright | s/man/horse/ |
| 19:02.13 | seanbright | you get the idea |
| 19:03.04 | rmudgett | gtjoseph: It's trying to be generic so on processors that support it it can get a speed advantage. |
| 19:07.23 | *** join/#asterisk-dev oej (~oej@h-205-12.A165.corp.bahnhof.se) |
| 19:07.23 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 19:25.49 | *** join/#asterisk-dev oej (~oej@h-205-12.A165.corp.bahnhof.se) |
| 19:25.49 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 19:37.27 | *** join/#asterisk-dev oej (~oej@h-205-12.A165.corp.bahnhof.se) |
| 19:37.27 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 19:40.02 | *** join/#asterisk-dev bford (d8cff501@gateway/web/freenode/ip.216.207.245.1) |
| 19:40.02 | *** mode/#asterisk-dev [+o bford] by ChanServ |