| 00:09.06 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-155-170.cablep.bezeqint.net) |
| 00:09.06 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 00:24.07 | *** join/#asterisk-dev newtonr (~newtonr@nat/digium/x-zmguxowyhuaxnbhv) |
| 00:31.45 | *** join/#asterisk-dev newtonr (~newtonr@nat/digium/x-plumqmlifdtsmzis) |
| 01:48.51 | *** join/#asterisk-dev otherwiseguy (~otherwise@69.23.116.90) |
| 02:20.33 | *** join/#asterisk-dev wudles (~wudles@gateway.secureinstrument.com) |
| 02:53.39 | *** join/#asterisk-dev sruffell (~sruffell@asterisk/the-kernel-guy/sruffell) |
| 02:53.39 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 19:01.08 | *** join/#asterisk-dev infobot (~infobot@rikers.org) |
| 19:01.08 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Tier 4 support is in #2,000 |
| 19:03.18 | russellb | Kobaz: if you're removing anything during a traversal, and traversal isn't LIST_TRAVERSE_SAFE, then it could be the problem. |
| 19:03.23 | russellb | mjordan++ |
| 19:03.31 | russellb | infobot: mjordan++ |
| 19:04.18 | b3nt_pin | did you mean ++mjordan? |
| 19:04.25 | mjordan | I'm a post increment kinda guy |
| 19:04.26 | b3nt_pin | mjordan++ creates a copy of the previous version. |
| 19:04.35 | Kobaz | russellb: saving the pointer to next beforehand will work fine |
| 19:05.01 | *** part/#asterisk-dev scgm11 (~Sebastian@r186-50-67-211.dialup.adsl.anteldata.net.uy) |
| 19:05.22 | b3nt_pin | owait, I am in the wrong chat room ! :) |
| 19:05.33 | russellb | b3nt_pin: this is C! :-p |
| 19:05.48 | russellb | and infobot doesn't know about ++foo i don't think |
| 19:05.48 | b3nt_pin | not really, I lurk here too... |
| 19:05.56 | kpfleming | overloads russellb's post-increment operator to call abort() |
| 19:06.00 | b3nt_pin | o i C |
| 19:06.20 | file | overloads the bacon reactor |
| 19:06.21 | russellb | hugs Python, where i++ isn't valid at all |
| 19:06.33 | mjordan | russellb: +1 |
| 19:06.47 | b3nt_pin | strongly cautions hugging large constricting snakes |
| 19:07.02 | file | hug a dog instead |
| 19:07.05 | russellb | yay dogs |
| 19:09.40 | Kobaz | yo dog |
| 19:09.43 | Kobaz | we heard you like dogs |
| 19:11.02 | kpfleming | we heard you like dogs, so we put some dogs on your dogs |
| 19:12.21 | Kobaz | heh |
| 19:20.40 | Kobaz | definitely a memory overwrite problem |
| 19:21.03 | kpfleming | get write-once memory. problem solved. |
| 19:21.06 | Kobaz | i'm printing an ast_str and it's printing something entirely different than what it should be |
| 19:21.07 | Kobaz | yeah |
| 19:21.08 | russellb | i usually like my dogs |
| 19:21.13 | russellb | today they drove me crazy so i left the house |
| 20:16.42 | Kobaz | hmmm |
| 20:16.46 | Kobaz | i just thought of something |
| 20:18.01 | Kobaz | in regards to ast_channel opacity... at the cost of slightly more overhead, and maintaining type safety, why not do struct ast_channel { void *data; } |
| 20:18.21 | Kobaz | and then you can pass around a struct ast_channel instead of a void pointer |
| 20:26.03 | russellb | you can still pass around a struct ast_channel *, even if all you have in the header is struct ast_channel; |
| 20:26.28 | Kobaz | you can pass it one |
| 20:26.36 | Kobaz | but the compiler wont check the type |
| 20:26.43 | kpfleming | yes it will |
| 20:26.52 | kpfleming | we do this all the time |
| 20:27.03 | Kobaz | well you can change the headers to be ast_channel i guess |
| 20:27.10 | Kobaz | but the internal function takes a void * |
| 20:27.28 | Kobaz | i was saying if the defs were all void * |
| 20:27.31 | Kobaz | then you dont get any checks |
| 20:27.45 | kpfleming | yep |
| 20:28.18 | russellb | why do the functions take a void * |
| 20:28.32 | Kobaz | so you don't monkey channel internals anymore |
| 20:28.43 | russellb | but you can do that without using void * |
| 20:29.31 | russellb | and it looks like main/channel_internal_api.c does that, so i'm not sure what you're saying |
| 20:29.48 | Kobaz | i remeber porting stuff to trunk and having to convert channel params to void * |
| 20:29.48 | Kobaz | lemme see |
| 20:30.06 | russellb | shouldn't have to ... |
| 20:30.53 | Kobaz | it was just a random thought based on what i remembered from a while ago, i'll have to find what i did |
| 20:35.51 | Kobaz | static void ast_channel_destructor(void *obj) |
| 20:35.51 | Kobaz | { |
| 20:35.51 | Kobaz | <PROTECTED> |
| 20:35.54 | Kobaz | stuff like that |
| 20:36.08 | kpfleming | that's an ao2 destructor function |
| 20:36.13 | Kobaz | oh okay |
| 20:36.18 | kpfleming | it has to take a void*, but you never call that function directly |
| 20:36.29 | kpfleming | which is why it's 'static' scope |
| 20:36.33 | Kobaz | yeah |
| 20:37.46 | Kobaz | i think i just copied a function as a template to work on something and i saw the void * and thought everything needed one now |
| 20:37.48 | Kobaz | okay never mind, heh |
| 21:32.24 | *** join/#asterisk-dev GameGamer43 (u5533@gateway/web/irccloud.com/x-wkyfusqnbbqqfpom) |
| 21:57.19 | *** join/#asterisk-dev WIMPy (~wimpy@e183095026.adsl.alicedsl.de) |
| 22:03.00 | *** join/#asterisk-dev mpe (~mpe@gate.ipvision.dk) |
| 22:11.16 | putnopvut | I just wrote a testsuite test, and it seems to be passing on the first try, which I don't trust. What makes me distrust the situation further is that there is nothing written to /tmp/asterisk-testsuite after the test is complete. Is the lack of logs in /tmp/asterisk-testsuite a sign things are not as they seem? |
| 22:16.25 | *** join/#asterisk-dev WIMPy (~wimpy@e183095026.adsl.alicedsl.de) |
| 22:31.22 | mjordan | putnopvut: yes |
| 22:31.38 | mjordan | a test 'passes' so long as it returns the right thing from its main. |
| 22:32.39 | putnopvut | mjordan: okay. I added some print statements into the test and none of them actually show up when I run the test, so I need to figure out what's wrong. It's probably something lame. |
| 22:33.22 | putnopvut | Ah, I see what's up. |
| 22:33.47 | putnopvut | I left out those crucial two lines at the bottom of all tests that tell the script to actually run main() |
| 22:34.02 | putnopvut | The script was essentially a no-op. |
| 22:34.23 | putnopvut | And now things are failing like I would expect :) |
| 22:35.07 | *** part/#asterisk-dev mpe (~mpe@gate.ipvision.dk) |
| 22:52.12 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-155-148.cablep.bezeqint.net) |
| 22:52.12 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |