March 8, 2019 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >
March 8, 2019 [00:04:46] *** dewf <dewf!~dewf@cpe-68-206-152-179.elp.res.rr.com> has joined #haiku[00:07:57] *** jjpx <jjpx!~jjpx@gateway/tor-sasl/jjpx> has quit IRC (Remote host closed the connection)[20:19:46] *** echelog <echelog!~echelog@108.61.103.42> has joined #haiku[20:19:46] *** ChanServ sets mode: +v echelog[20:21:29] <k_> I have uploaded the screenshot of Haiku on imagebin and added its link to Outreachy[20:21:44] <Not-8d6a> [haiku/infrastructure] kallisti5 pushed 1 commit to master [+1/-1/±1] https://git.io/fhhGa[20:21:45] <Not-8d6a> [haiku/infrastructure] kallisti5 b158fa3 - k8s/ingress: Properly leverage load balancer[20:22:11] <k_> kallisti5: will it work?[20:23:13] *** Dariost <Dariost!~Dariost@151.20.212.91> has joined #haiku[20:23:43] <PulkoMandy> k_: you need to submit the changes to http://review.haiku-os.org[20:23:56] <PulkoMandy> when you have done that we will give you the appropriate link[20:24:55] <k_> PulkoMandy: But there I won't need the screenshot, right?[20:25:11] <PulkoMandy> yes, the code change is more important than the screenshot[20:25:30] <PulkoMandy> as long as you made sure you are able to compile and test things, you are fine[20:25:44] <k_> okay, thanks[20:25:46] <PulkoMandy> the screenshot is just a proof for that, but we trust you[20:26:05] <PulkoMandy> if you get into problems with the building part, you will have to solve them anyway, so why would you lie about it :)[20:26:24] <k_> :D[20:27:32] <k_> How do I add my file to review.haiku-os.org ?[20:28:09] *** gouchi <gouchi!~gouchi@unaffiliated/gouchi> has joined #haiku[20:29:01] <k_> I'm familiar with github, but I couldn't push the file on gerrit[20:29:07] *** Pentode <Pentode!~pentode@mobile-166-170-23-158.mycingular.net> has joined #haiku[20:30:09] <jjpx> k_, https://www.haiku-os.org/guides/building/get-source-git/[20:31:09] *** Mazon <Mazon!~Mazon@37.205.127.168> has joined #haiku[20:31:51] <jjpx> also https://dev.haiku-os.org/wiki/CodingGuidelines/SubmittingPatches[20:32:02] *** superprower <superprower!~superprow@188.130.155.151> has joined #haiku[20:32:11] <jjpx> they are almost the same, but not quite[20:32:15] *** SpiceGuid <SpiceGuid!~chatzilla@alyon-158-1-43-204.w90-53.abo.wanadoo.fr> has quit IRC (Quit: ChatZilla 0.9.93 [SeaMonkey 2.49.4/20180713174829])[20:33:33] *** k_ <k_!b64b2d01@gateway/web/freenode/ip.182.75.45.1> has quit IRC (Ping timeout: 256 seconds)[20:40:35] *** Maturion <Maturion!~Maturion@p200300EDE70E2100911348D073D64DB7.dip0.t-ipconnect.de> has joined #haiku[20:45:42] *** Psud0nym <Psud0nym!~Psud0nym@2601:cc:8100:6244:a8e9:997:3abe:635a> has joined #haiku[20:48:57] <superprower> PulkoMandy: hey! I've compared some code between older version and master, but I don't really see a lot of stuff actually setting it, literally one place in the entire driver that changes frequencies that hasn't change in the past several years. So I think I shouldn't be looking at frequencies themselves, but for something else? Divisors, maybe?[20:49:22] <superprower> * it's about PLL settings in intel_extreme driver, forgot to mention that[20:49:39] *** Psud0nym <Psud0nym!~Psud0nym@2601:cc:8100:6244:a8e9:997:3abe:635a> has quit IRC (Client Quit)[20:50:52] *** HAIKU-irker008 <HAIKU-irker008!~irker@walter.haiku-os.org> has joined #haiku[20:50:52] <HAIKU-irker008> haiku.master: waddlesplash * hrev52973 [1 commit] https://git.haiku-os.org/haiku/log/?qt=range&q=ba56d0a513ed+%5E901c3d44b047[20:50:52] <HAIKU-irker008> ba56d0a513ed: pthread: add pthread_getattr_np.[20:50:52] *** ChanServ sets mode: +v HAIKU-irker008[20:52:46] *** Amigori <Amigori!~amigori@209-112-213-52-radius.dynamic.acsalaska.net> has joined #haiku[20:54:13] <PulkoMandy> superprower: yes, the frequency we need is computed by the GTF, actually I think this is shared between all drivers[20:54:30] <PulkoMandy> and the divisor constraints were changed a bit but I think they are all right now[20:54:58] <PulkoMandy> again, it is very unlikely that your problem is PLLs. That kallisti5 once had a bad experience with them does not mean they are the source of all problems in the driver...[20:55:39] <superprower> Uh, "again"? oops. I think I misunderstood you last time.[20:56:39] <PulkoMandy> do you have a syslog?[20:57:13] *** Amigori <Amigori!~amigori@209-112-213-52-radius.dynamic.acsalaska.net> has quit IRC (Ping timeout: 246 seconds)[20:57:14] <PulkoMandy> unless the logs about the frequency there is obviously wrong (like, "trying 130MHz, found 86MHz, that should do"), I don't think a PLL problem is likely[20:57:39] <PulkoMandy> you could rather check all the pipe and FDI link setup, which is half finished and has changed a lot more between chipset generations[20:58:16] <PulkoMandy> what hardware are you working with anyway? And is there a ticket tracking your problems?[20:59:21] *** axeld <axeld!~Thunderbi@89.233.79.10> has quit IRC (Remote host closed the connection)[21:02:40] <superprower> Was just going to try register dump and reading syslog. I'm working with GMA 900 (i915GM). Just found a related bug - #14280. It's interesting because guy there says about "zero active displays were found" error, which I got when I tried some revision that just hanged on a loading screen instead of just being a black screen. I'll try confirming that this isn't happening in older version.[21:03:56] <PulkoMandy> ok, so that's for older hardware. Great, you have better docs for it, and no FDI :)[21:04:18] <superprower> Yeeeah, about docs...[21:04:41] <superprower> Docs at 01.org start juuust from the next generation after mine.[21:05:52] *** illwieckz <illwieckz!~illwieckz@unvanquished/developer/illwieckz> has quit IRC (Read error: Connection reset by peer)[21:05:57] *** illwieckz_ <illwieckz_!~illwieckz@unvanquished/developer/illwieckz> has joined #haiku[21:07:07] <superprower> Or even two generations, i.e. chipset 965 is the oldest there.[21:09:19] <PulkoMandy> according to https://en.wikipedia.org/wiki/List_of_Intel_graphics_processing_units you have GMA900[21:09:41] <PulkoMandy> and according to https://en.wikipedia.org/wiki/Intel_GMA GMA900 and GMA950 are both 3rd generation devices, so they should be similar[21:09:49] <PulkoMandy> but maybe not identical, indeed...[21:10:08] *** illwieckz_ <illwieckz_!~illwieckz@unvanquished/developer/illwieckz> has quit IRC (Client Quit)[21:10:34] *** illwieckz <illwieckz!~illwieckz@unvanquished/developer/illwieckz> has joined #haiku[21:12:20] <PulkoMandy> volume 3 of the specs is the most relevant for our purposes[21:12:28] <PulkoMandy> (as long as you just want modesetting and no 3D acceleration)[21:15:33] <superprower> I'll keep that in mind, thanks[21:15:45] *** Begas_VBox <Begas_VBox!~Begasus@d54C05BEA.access.telenet.be> has quit IRC (Quit: Vision[0.10.3]: i've been blurred!)[21:16:23] <Begasus> heading down, g'night peeps[21:16:29] *** Begasus <Begasus!~begasus@ptr-4qba4ea1g2xif3foi77.18120a2.ip6.access.telenet.be> has quit IRC (Quit: Ex-Chat)[21:17:06] *** BlueSky76 <BlueSky76!~mutscho@80-108-11-4.cable.dynamic.surfer.at> has left #haiku[21:17:57] <superprower> So tomorrow I'll try comparing register dumps and syslogs and I'll try comparing two versions in regards for Pipe and FDI link setup. Except FDI didn't even appear until Gen 5, according to wikipedia?[21:18:43] *** axeld <axeld!~Thunderbi@89.233.79.10> has joined #haiku[21:18:43] *** ChanServ sets mode: +o axeld[21:19:30] <PulkoMandy> yes, no FDI for older devices so that's one thing less to worry about[21:19:54] <PulkoMandy> you have just two display pipes, and the two ports associated with them (usually LVDS and VGA or HDMI for laptops)[21:21:01] <superprower> Is FDI compatibility reported by hardware, or is there a possibility that driver may attempt to set it up for incompatible device? Just curious.[21:21:09] *** bos20k <bos20k!~bos20k@ool-4579e359.dyn.optonline.net> has quit IRC (Quit: Bye.)[21:22:06] *** Lelldorin1 <Lelldorin1!~lelldorin@x4d0150d6.dyn.telefonica.de> has joined #haiku[21:22:16] <PulkoMandy> the hardware does not report anything about features that would be introduced only 10 years later, of course. The engineer didn't even know FDI would exist someday[21:22:38] <PulkoMandy> we work from the PCI ID to know which hardware we are dealing with[21:22:46] <PulkoMandy> and there are if() all over the place in the driver to do the right thing[21:23:20] <PulkoMandy> it's quite possible at some place we just forgot to put the oldest card as "same as GMA950" in one of the if() clauses[21:28:35] <jjpx> the ideal case would be a dedicated driver with no if's at all?[21:28:51] <PulkoMandy> no, there is still a lot of shared code[21:29:04] <PulkoMandy> nd it wouldn't be 1 dedicated driver, it would be a dozen of them...[21:30:04] <jjpx> that explains the size of windows drivers downloads[21:30:17] <waddlesplash> no, windows drivers are just stupid[21:30:42] <waddlesplash> Linux drivers are massive (probably millions or LoC per graphics family) but compiled they are still < 10MB each[21:30:52] <waddlesplash> heck, WebKit's repo is larger than 5GB, but WebKit compiled is only 20MB[21:30:56] <PulkoMandy> 10MB is still a crazy lot of code...[21:31:06] <waddlesplash> indeed it is[21:31:14] <PulkoMandy> 5GB including tests, website, and the git history[21:31:20] <waddlesplash> right[21:31:34] <PulkoMandy> it's probably only a few hundred megabytes if you keep just the sources and only 1 version of it[21:33:05] <waddlesplash> ok, so now that XHCI almost works properly, I guess I should attempt isochronous transfers and XHCI-in-KDL[21:33:27] <superprower> What would last thing do?[21:33:40] <waddlesplash> looking at the EHCI driver, it is actually shocking how much easier it is to do isochronous transfers in XHCI than EHCI[21:34:03] <waddlesplash> partially because the hardware does all the heavy lifting for us, instead of the driver under EHCI[21:34:14] <DHowett> it's crazy to think of a live kernel debugger on a live system where you need a _pretty good_ amount of the kernel working to do the debugging unless you need to debug the part that is needed to to the debugging[21:34:31] <waddlesplash> DHowett: lol[21:34:41] <waddlesplash> actually we really don't need much of the kernel running in KDL[21:34:48] <PulkoMandy> there were serial ports once...[21:35:01] <PulkoMandy> also EHCI has a special "debug port" feature but we never wrote a driver for that[21:35:02] <waddlesplash> that's why I have to write stuff in XHCI driver to get that working in KDL, because we can't do interrupts or locking[21:35:09] <waddlesplash> PulkoMandy: XHCI does too[21:35:15] <PulkoMandy> and it is not as simple as it should have so we may as well get keyboards working :>[21:35:20] <jjpx> serial ports are still used to configure cisco routers[21:35:26] <jjpx> we used rs232[21:35:37] <DHowett> i know of a debug board for the PC i'm currently using that would make KD stuff easier at least for Windows ...[21:35:40] <DHowett> i should try to source one[21:35:42] <superprower> Serial ports are still used to debug OSes, as far as I know[21:35:51] <waddlesplash> superprower: we use them, indeed[21:35:51] <PulkoMandy> when you have one, yes[21:36:06] <DHowett> jjpx: i was somewhat appalled as I recently had to construct an 8p8c rs232 adapter for a machine i was building[21:36:20] <waddlesplash> but adding support to the XHCI driver so we can use it in the kernel debugger would make it possible to e.g. use USB3-attached keyboards[21:37:03] <superprower> By the way got notification that my new serial to usb that isn't based on cheap adapter is at post office, so it'll make debugging a bit easier for me, I guess[21:37:36] <PulkoMandy> serial to USB won't help in KDL or for grabbing syslogs...[21:37:52] <PulkoMandy> you need a real serial port, on the motherboard, or as a PCI or ExpressCard device[21:38:16] <superprower> I know[21:38:19] <superprower> It's for recieving end[21:38:25] <PulkoMandy> ah, ok :)[21:38:30] <PulkoMandy> yes that side works[21:38:50] <superprower> It's just that cheap CH340 adapter was freaking out when I tried actually reading something from Haiku[21:38:58] <waddlesplash> at this rate, XHCI will work better than any other USB hardware :P[21:39:08] <waddlesplash> someone just needs to fix that last "can't boot from USB3 on NEC hardware" bug[21:39:12] <PulkoMandy> mh... do we even have a driver for CH340?[21:39:32] <PulkoMandy> and... did they translate the datasheet for this one or is it chinese only?[21:40:37] <waddlesplash> PulkoMandy: https://sparks.gogo.co.nz/ch340.html -- non GPL (??) Linux driver source code[21:41:02] <PulkoMandy> waddlesplash: I worked with WCH chips for other projects[21:41:06] <waddlesplash> actually there is no license in the file, lol[21:41:16] <PulkoMandy> they have docs on their website but usually in chinese[21:41:26] <waddlesplash> well, the driver comments are in English[21:41:28] <PulkoMandy> yes, not surprised, probably it's derived from their sample files[21:41:39] <PulkoMandy> chinese have a different view on copyright[21:41:49] <PulkoMandy> aka "this is software, we don't sell it, we don't care"[21:42:41] <PulkoMandy> http://wch.cn/search?t=all&q=CH340[21:42:42] <brj> lol[21:42:49] <PulkoMandy> datasheet, examples, sourcecode for drivers, all you need[21:43:54] <jjpx> we have some chinese contributors[21:44:01] <jjpx> or not?[21:44:05] <PulkoMandy> they also have a forum, mostly chinese but they will reply to english questions[21:44:21] <PulkoMandy> not much, and I think mostly in the translation team[21:44:40] <PulkoMandy> not sure they are technical enough to translate these docs to readable english[21:44:50] <PulkoMandy> last time I needed some docs from there I just used Google Translate[21:46:39] <jjpx> do they use chinese characters to name variables in source?[21:47:04] <PulkoMandy> no, it is not allowed in C[21:48:05] *** Lelldorin1 <Lelldorin1!~lelldorin@x4d0150d6.dyn.telefonica.de> has quit IRC (Quit: Vision[0.10.3]: i've been blurred!)[21:49:33] *** Maturion <Maturion!~Maturion@p200300EDE70E2100911348D073D64DB7.dip0.t-ipconnect.de> has quit IRC (Quit: Leaving)[21:53:04] <waddlesplash> they may use pinyin though[21:53:10] <waddlesplash> but it appears they do not[21:59:09] *** axeld <axeld!~Thunderbi@89.233.79.10> has quit IRC (Remote host closed the connection)[22:02:15] <DepositePirate> Finally I've cracked how to convert 24 bit color to 16 color irgb without any quantization or k-means or whatever bullshit!!!![22:02:17] *** axeld <axeld!~Thunderbi@89.233.79.10> has joined #haiku[22:02:17] *** ChanServ sets mode: +o axeld[22:02:48] <B2IT> (AGMS) By bit shifting?[22:03:04] <DepositePirate> nope[22:03:14] <B2IT> (AGMS) Really huge table lookup?[22:03:33] <DepositePirate> well you have to do that too, but you simply have to calculate the brightness bit separately.[22:04:38] <B2IT> (AGMS) So, is this for converting graphics to CGA text mode, where there are 16 colours and brightness (or was it 8 colours)?[22:04:44] <DepositePirate> yes, right[22:05:04] <DepositePirate> CGA text mode, which is what consoles still use today.[22:05:13] <B2IT> (AGMS) Hmmm, something for a boot screen perhaps?[22:05:32] <DepositePirate> Nope for displaying XPM images in the console :)[22:05:53] <B2IT> (AGMS) Ah, quick preview without needing a windowing system etc![22:06:11] <DepositePirate> No, specially crafted ones to print logos :)[22:06:45] <DepositePirate> But which you'll be able to edit in your favorite imaging program and then compile in or load[22:07:47] <DepositePirate> Which is why I don't need some fancy quantization or k-means technique[22:07:58] <B2IT> (AGMS) Could also display it on a DEC VT241 terminal :-)[22:08:45] <DepositePirate> Well hopefully the thing will work with anything that can compile C and has a screen :)[22:08:58] *** Amigori <Amigori!~amigori@209-112-213-52-radius.dynamic.acsalaska.net> has joined #haiku[22:10:59] <B2IT> (AGMS) Got sidetracked by https://vt100.net/dec/vt_history[22:15:07] *** brj <brj!~brj@123.201.163.64> has quit IRC (Ping timeout: 240 seconds)[22:20:38] *** leemon <leemon!32075d1c@gateway/web/freenode/ip.50.7.93.28> has joined #haiku[22:25:28] *** Vidrep_64 <Vidrep_64!~vision@d75-156-156-137.abhsia.telus.net> has joined #haiku[22:28:42] *** axeld <axeld!~Thunderbi@89.233.79.10> has quit IRC (Remote host closed the connection)[22:32:53] <leemon> hi, I was looking into crypto kit rand_str function, can we use some source of entropy? I can't find how srand() is implemented maybe we should think about more or less random random_number_generator?[22:35:38] *** leemon_ <leemon_!32075d1c@gateway/web/freenode/ip.50.7.93.28> has joined #haiku[22:37:39] *** leemon <leemon!32075d1c@gateway/web/freenode/ip.50.7.93.28> has quit IRC (Ping timeout: 256 seconds)[22:40:10] <Not-8d6a> [haiku/infrastructure] kallisti5 pushed 1 commit to master [+0/-0/±1] https://git.io/fhhc4[22:40:12] <Not-8d6a> [haiku/infrastructure] kallisti5 8a9bd56 - k8s: add -persist to ensure volumes stick around on delete[22:43:11] *** leemon_ <leemon_!32075d1c@gateway/web/freenode/ip.50.7.93.28> has quit IRC (Quit: Page closed)[22:46:15] *** axeld <axeld!~Thunderbi@89.233.79.10> has joined #haiku[22:46:15] *** ChanServ sets mode: +o axeld[22:52:08] <bbjimmy> I am unable to create a pull request on haikuports. no browser on haiku works to actually press the create pull request button and do anything.[22:55:00] <bbjimmy> Is there a way to do this without using a web browser?[22:58:40] *** axeld <axeld!~Thunderbi@89.233.79.10> has quit IRC (Remote host closed the connection)[23:01:10] *** axeld <axeld!~Thunderbi@89.233.79.10> has joined #haiku[23:01:11] *** ChanServ sets mode: +o axeld[23:07:00] <jjpx> bbjimmy, you can push your changes via git and thenm use other os to create the PR[23:07:03] *** Amigori <Amigori!~amigori@209-112-213-52-radius.dynamic.acsalaska.net> has quit IRC (Remote host closed the connection)[23:07:30] <jjpx> even a smartphone[23:07:33] *** Amigori <Amigori!~amigori@209-112-213-52-radius.dynamic.acsalaska.net> has joined #haiku[23:08:16] *** Dariost <Dariost!~Dariost@151.20.212.91> has quit IRC (Quit: Dariost)[23:08:37] *** Dariost <Dariost!~Dariost@151.20.212.91> has joined #haiku[23:12:59] <HAIKU-irker008> haiku.master: korli * hrev52974 [1 commit] https://git.haiku-os.org/haiku/log/?qt=range&q=9ba45ef9246e+%5Eba56d0a513ed[23:13:00] <HAIKU-irker008> 9ba45ef9246e: Update libroot stubs.[23:14:32] *** Dariost <Dariost!~Dariost@151.20.212.91> has quit IRC (Quit: Dariost)[23:14:45] *** Dariost <Dariost!~Dariost@151.20.212.91> has joined #haiku[23:17:57] <bbjimmy> jjpx I have no other OIS[23:18:01] <bbjimmy> OS[23:18:04] *** Amigori <Amigori!~amigori@209-112-213-52-radius.dynamic.acsalaska.net> has quit IRC (Remote host closed the connection)[23:18:15] <bbjimmy> I have no smartphone[23:24:37] <jjpx> no ideas (other than borrow a phone)[23:25:34] *** Amigori <Amigori!~amigori@209-112-213-52-radius.dynamic.acsalaska.net> has joined #haiku[23:26:18] *** Megaf <Megaf!~Megaf@unaffiliated/megaf> has quit IRC (Quit: Leaving)[23:28:48] *** axeld <axeld!~Thunderbi@89.233.79.10> has quit IRC (Remote host closed the connection)[23:29:10] *** axeld <axeld!~Thunderbi@89.233.79.10> has joined #haiku[23:29:10] *** ChanServ sets mode: +o axeld[23:33:39] *** ClaudioM <ClaudioM!~ClaudioM@c-73-46-231-176.hsd1.fl.comcast.net> has joined #haiku[23:34:36] *** bbjimmy <bbjimmy!~bbjimmy@host-67-44-193-24.hnremote.net> has quit IRC (Quit: Vision[0.10.3]: i've been blurred!)[23:37:24] *** embden <embden!~embden@185.202.212.202> has quit IRC (Ping timeout: 252 seconds)[23:39:09] *** gouchi <gouchi!~gouchi@unaffiliated/gouchi> has quit IRC (Remote host closed the connection)[23:44:38] *** pemdp <pemdp!~pedro@a95-94-254-121.cpe.netcabo.pt> has joined #haiku[23:47:00] *** AD_MOS2 <AD_MOS2!~JohnN@176.255.158.95> has quit IRC (Ping timeout: 250 seconds)[23:48:47] *** hbelusca <hbelusca!~hbelusca@reactos/developer/hbelusca> has joined #haiku[23:54:24] *** axeld <axeld!~Thunderbi@89.233.79.10> has quit IRC (Quit: axeld)[23:56:26] *** Amigori <Amigori!~amigori@209-112-213-52-radius.dynamic.acsalaska.net> has quit IRC (Remote host closed the connection)