| brolin_empey | CatButts: My former colleague said he thinks the model of Windows Mobile smartphone he had is the HTC Touch: https://en.wikipedia.org/wiki/HTC_Touch | 06:58 |
|---|---|---|
| CatButts | https://elixir.bootlin.com/linux/v2.6.29/source/include/linux/netfilter_ipv4/ipt_MARK.h | 12:59 |
| CatButts | https://elixir.bootlin.com/linux/v2.6.29/source/include/linux/netfilter_ipv4/ipt_mark.h | 12:59 |
| CatButts | case sensitive filenames, nice | 13:00 |
| CatButts | renaming them(and the includes) is easy | 13:02 |
| KotCzarny | well, you know those are 2 different files, right? | 13:02 |
| CatButts | but problem comes when you distribute sourcecode relying on modified ones | 13:02 |
| CatButts | yes | 13:02 |
| KotCzarny | just use ntfs | 13:03 |
| KotCzarny | instead of fat | 13:03 |
| CatButts | I am using ntfs | 13:03 |
| KotCzarny | ntfs is case sensitive last time i checked | 13:03 |
| CatButts | file winapi needs to explicitly request case sensitivity | 13:03 |
| CatButts | guess how often that happens :p | 13:04 |
| CatButts | also, you have registry setting negating that request and giving you case insensitive anyway | 13:04 |
| CatButts | also, windows 10 added a mechanism for setting per-folder(non recursive) case sensitivity | 13:05 |
| CatButts | also, I don't care about windows 10 | 13:05 |
| bencoh | honestly I wouldn't even try at that point | 13:07 |
| CatButts | why, because a few headers were badly named? | 13:08 |
| bencoh | I had the very same issue back when I tried to build a linux kernel from macosx (hfs was mostly case-insensitive back then) | 13:08 |
| bencoh | because I suspect there will be more | 13:08 |
| bencoh | and there's nothing to gain from the experience | 13:09 |
| CatButts | http://cat.butt.care:3434/these_ones.png | 13:09 |
| bencoh | oh and, you'll have to bang your head against the whole kernel build process | 13:09 |
| bencoh | I mean, makefile defines relying on script resulsts | 13:09 |
| CatButts | hmmm, ouch | 13:10 |
| bencoh | I dunno how many of those there were back in 2009 | 13:10 |
| bencoh | but honestly I wouldn't even think of it in 2020 | 13:11 |
| CatButts | thankfully, I like my things simple | 13:11 |
| CatButts | though, will be problem when compiling someone else's application | 13:11 |
| CatButts | if gcc has defines for telling which host platform it is running on, then that is my solution to renamed header | 13:13 |
| bencoh | to be honest it looks better than I thought after looking at the makefile | 13:13 |
| bencoh | but you'll still have work to do there | 13:13 |
| CatButts | > and there's nothing to gain from the experience | 13:16 |
| CatButts | I get to be lazy and comfortable afterward | 13:16 |
| CatButts | but you know what they say, about lazy people ending up running more | 13:16 |
| CatButts | haha | 13:17 |
| KotCzarny | you wouldnt be doing it if lazy | 13:17 |
| CatButts | are there any parameters I should care of when operating ARM gcc crosscompiler, compared to gcc for PC? | 14:57 |
| KotCzarny | on crosscompilation yes | 14:57 |
| KotCzarny | when native it finds them out itself | 14:57 |
| CatButts | say, for a hello_world.c | 14:57 |
| CatButts | console printf | 14:58 |
| KotCzarny | compile, test | 14:58 |
| KotCzarny | if it doesnt work, then yes | 14:58 |
| CatButts | I'll find out after putting together bastard GCC | 14:58 |
| CatButts | but I have a feeling it won't work | 14:58 |
| bencoh | for basic stuff it should work almost out of the box | 14:59 |
| bencoh | assuming the toolchain is properly set up, and sysroot is found in expected place | 14:59 |
| bencoh | (when build kernels you don't even need any sysroot) | 15:00 |
| bencoh | building* | 15:00 |
| CatButts | applications | 15:00 |
| CatButts | sysroot is GCC's base folder, yes? | 15:01 |
| KotCzarny | dont forget to pass ARCH=arm when building kernel | 15:01 |
| CatButts | ah yeah, one more thing to ask, default functions | 15:04 |
| bencoh | sysroot is more than that | 15:04 |
| CatButts | on windows, there are 3 of them | 15:04 |
| bencoh | it contains target libs/headers as well | 15:05 |
| CatButts | int main() for console, WinMain() for GUI applications, DllMain() for dlls | 15:05 |
| CatButts | what's it like on linux? | 15:05 |
| KotCzarny | main() | 15:05 |
| CatButts | just the one | 15:06 |
| KotCzarny | and for gui it just depends what toolkit you use | 15:06 |
| KotCzarny | most likely qt5 | 15:06 |
| CatButts | yeah no | 15:06 |
| CatButts | the compiler name is no longer gcc.exe/gcc | 15:19 |
| CatButts | http://cat.butt.care:3434/ohwellitried.png | 15:31 |
| CatButts | I am close though, as I have hello world printf running, from unaltered GCC 4.9 | 15:48 |
| CatButts | and now transplant works | 16:57 |
| CatButts | apparently I am only partially done as the bulk of headers and libs are with scratchbox | 18:24 |
| CatButts | how does SB interact with GCC, regarding the sysroot | 18:25 |
| bencoh | it's ... pretty awful tbh | 18:27 |
| bencoh | SB replaces the "default" (ARM-native) compiler with the toolchain's ARM crosscompiler | 18:28 |
| CatButts | oh, it duplicates gcc | 18:31 |
| CatButts | all this time, I've been toying with default gcc headers and libs | 18:31 |
| KotCzarny | sb inteacts with gcc.. a lot | 18:33 |
| KotCzarny | ;) | 18:33 |
| bencoh | it's not just gcc by the way | 18:35 |
| CatButts | soo, SB is an overcomplicated monster that keeps things simple, and attempting a 2nd transplant over gcc folders might result in patient death | 18:38 |
| CatButts | I assume that if a specific header/lib is found both in gcc and scratchbox, the one in SB will take precedence, yes? | 21:45 |
| KotCzarny | ask gcc about paths | 22:14 |
| * enyc meows | 23:02 | |
| * CatButts farts | 23:09 | |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
