00:17.36 | *** join/#htc-linux XXXXXXXX (~XXXXXXXX@74-139-20-5.dhcp.insightbb.com) |
00:19.07 | XXXXXXXX | HI ANYONE HERE |
00:39.47 | *** join/#htc-linux eXtremo (~eXtremo@unaffiliated/extremo) |
01:32.56 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
02:02.47 | *** part/#htc-linux langpoo (~zz@109.205.253.11) |
02:12.20 | *** part/#htc-linux Cotulla (~myfakemai@109.205.253.11) |
02:29.08 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
02:41.21 | detule | jonpry, qcom offer you a job yet? |
02:43.18 | jonpry | not yet detule |
02:43.45 | jonpry | this talk of releasing the HVM is interesting |
02:44.05 | jonpry | i have serious doubts it will happen though |
02:44.39 | jonpry | even if the engineer is serious, it's probably the same idea they had years ago |
02:44.48 | detule | must be hard working there, if everything about the lawyers there is true |
02:45.26 | jonpry | stupid lawyers |
02:45.40 | jonpry | this SMP is driving me nuts |
02:46.23 | detule | fork() still failing? |
02:46.24 | jonpry | insane things start happening |
02:46.32 | jonpry | fork() is fixed |
02:46.45 | jonpry | afaict it's 100% with a single processor |
02:46.57 | jonpry | although Cotulla talks of some kind of slowdown |
02:47.44 | detule | so how is smp failing? |
02:48.03 | jonpry | that isn't well understood |
02:48.14 | jonpry | things just stop making sense |
02:48.48 | jonpry | at some point it will either hang, have strange behavior with the TLB, or start causing double faults |
02:49.02 | jonpry | i think i understand the hang though |
02:49.37 | detule | how does smp come into play are you guys onlin-ing another hexagon? |
02:50.03 | jonpry | i have it bringing up the second "core" |
02:50.24 | detule | what second core? |
02:50.38 | jonpry | so v2 is 6, v3 is quad |
02:50.40 | jonpry | i think |
02:50.54 | jonpry | no testing has been done beyond 2. but that is the word on the street |
02:52.02 | jonpry | it's not like a real core |
02:52.19 | jonpry | it's more like hyperthreading |
02:52.48 | jonpry | except not |
02:53.34 | jonpry | like on 6 core. if a single thread is running. it only will issue say 600mhz/6 = 100M instructions / second |
02:54.09 | jonpry | the pipeline is 6 clocks long. so the delay causes instructions to always appear to take 1 clock cycle |
02:54.16 | jonpry | with no delays and crud |
02:55.47 | detule | is there any documentation on how they are virtualizing the core |
02:56.11 | jonpry | virtualizing? |
02:56.21 | detule | how they are presenting these 6 cores |
02:56.41 | jonpry | there isn't really any documentation for anything |
02:56.46 | jonpry | but we know how to start it |
02:57.28 | jonpry | every core just has it's own register file |
02:57.41 | jonpry | and 1/6 clock cycles it gets access to the ALU |
02:57.50 | detule | so the linux kernel actually sees 6 processors? |
02:57.55 | jonpry | yes |
02:57.58 | jonpry | 2 for now |
03:01.44 | jonpry | processor <0x09> : 0 |
03:01.45 | jonpry | model name <0x09> : QDSP6V3 |
03:01.47 | jonpry | BogoMips <0x09> : 200.00 |
03:01.47 | jonpry | processor <0x09> : 1 |
03:01.47 | jonpry | model name <0x09> : QDSP6V3 |
03:01.47 | jonpry | BogoMips <0x09> : 200.00 |
03:02.09 | jonpry | cat /proc/cpuinfo :) |
03:02.10 | detule | nice |
03:02.23 | detule | console working? |
03:02.28 | detule | "Include mistic zuzu logger here" ? :) |
03:02.46 | jonpry | mine is ram_console |
03:03.32 | jonpry | i don't have zuzu |
03:15.14 | detule | have you heard of this MobiCore business |
03:16.08 | jonpry | no |
03:17.29 | detule | MobiCore is an operating system being shipped with TZBSP |
03:17.29 | detule | on msm chipsets. MobiCore consists of several components in |
03:17.29 | detule | the secure world(TrustZone) and non-secure world(linux |
03:17.29 | detule | kernel, Android user space). The MobiCore driver |
03:17.29 | detule | communicates with the MobiCore kernel that exists in |
03:17.32 | detule | TrustZone. |
03:17.50 | jonpry | hmm |
03:17.56 | jonpry | can we get it to load code? |
03:18.16 | detule | http://code.google.com/p/tee-mobicore-driver/ <--- source for the daemon |
03:18.27 | detule | i actually saw commits |
03:18.33 | detule | in the caf trees |
03:18.35 | detule | today |
03:19.05 | jonpry | like do the mobicore apps have to be signed somehow? |
03:38.26 | detule | i doubt this thing will make anyone's life easier |
03:39.05 | jonpry | could be |
03:39.24 | jonpry | i wonder what exactly a "high security operating system" is |
03:40.25 | detule | https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&ved=0CEIQFjAC&url=http%3A%2F%2Fwww.gi-de.com%2Fgd_media%2Fmedia%2Fdocuments%2Fcomplementary_material%2Fevents_1%2F04_STE_CARTES__Demo_Presentation.pdf&ei=_6E-UZ3cL4XL0AGDjYDADQ&usg=AFQjCNFFazRZtP606PQRpALzGZHEuvs4PQ&sig2=t0MPFNKYixT4a2VrXHN1zQ&bvm=bv.43287494,d.dmQ |
03:41.08 | detule | i don't know if this thing is actually loaded on current msm |
03:41.16 | detule | except that this doc almost makes it sound like it is |
03:41.16 | detule | https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&ved=0CEcQFjAD&url=http%3A%2F%2Ftee-mobicore-driver.googlecode.com%2Fgit-history%2F057a4f07d5e23916c1f353abe63d2afb5153dccd%2FV008%2FRelease_Notes.pdf&ei=_6E-UZ3cL4XL0AGDjYDADQ&usg=AFQjCNGOOsoKJCJeJhKtmpDCIVoiy9GNSg&sig2=JTRhK_onJTEAm_V_gy8XTw&bvm=bv.43287494,d.dmQ |
03:44.00 | detule | g&d i think these guys are a no-joke outfit ... like they design passport security features for governments |
03:45.07 | jonpry | looks pretty serious |
03:45.43 | jonpry | even if i could make a trustlet, it seems like breaking that mobicore could be a real hassle |
03:48.45 | jonpry | the easiest way to access qdsp on 8960 imho, is some kind of exploit in the q6 firmware itself |
03:49.05 | jonpry | doesn't seem like it would be written with a lot of security in mind |
05:13.27 | *** join/#htc-linux jonpry (~jon@user-0c6tmn0.cable.mindspring.com) |
09:22.17 | *** join/#htc-linux apt (~apt@rikers.org) |
09:22.17 | *** topic/#htc-linux is Welcome to the HTC Linux project | Community portal & WiKi http://htc-linux.org | For IRC logs, HaRET & kernel mailing lists etc. see http://htc-linux.org/wiki/index.php?title=Contact | The htc-linux.org project is not affiliated with the HTC Corporation | This channel is for development purposes - Join #htc-linux-chat for offtopic |
10:20.38 | gauner1986 | fakker: when is it scheduled to ship? |
10:42.53 | fakker | it's due 15th |
10:42.55 | fakker | so this Friday |
10:43.04 | fakker | i'd hope it'll ship Thursday latest... |
11:34.05 | *** join/#htc-linux surgex (~surge@pool-98-118-154-23.bflony.fios.verizon.net) |
11:36.54 | *** join/#htc-linux paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
11:37.34 | *** join/#htc-linux phh_ (~quassel@137.194.15.151) |
11:40.26 | gauner1986 | fakker: hah.. mine should ship on 18th/19th |
11:42.36 | *** join/#htc-linux bartman (~bart@2607:f2c0:f00e:700::dd) |
12:11.54 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
12:29.55 | *** join/#htc-linux paulk-desktop (~paulk@82.233.88.171) |
13:50.02 | *** join/#htc-linux helicopter88 (~helicopte@95.234.116.121) |
13:50.04 | *** join/#htc-linux eXtremo (~eXtremo@unaffiliated/extremo) |
13:57.16 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
14:06.04 | *** join/#htc-linux bartman (~bart@2607:f2c0:f00e:700::dd) |
14:32.17 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
15:14.01 | *** join/#htc-linux Marc (~quassel@phpbb/modifications/marc1706) |
15:37.45 | *** join/#htc-linux bartman (~bart@2607:f2c0:f00e:700::dd) |
15:54.18 | *** join/#htc-linux Mavy-bnc (mavfree@nlck1m.ckgaming.co.uk) |
15:59.07 | *** join/#htc-linux Ondalf (~ondalf@cable-roi-50ddf8-39.dhcp.inet.fi) |
16:09.53 | *** join/#htc-linux langpoo (~zz@109.205.253.11) |
16:09.53 | *** join/#htc-linux Cotulla (~myfakemai@109.205.253.11) |
16:12.51 | *** join/#htc-linux kiozen (~kiozen@p578a42db.dip0.t-ipconnect.de) |
16:16.09 | Cotulla | detule I heard u fixed our issue |
16:16.14 | Cotulla | is it true? |
16:16.31 | Cotulla | oh u are d3tul3 now |
16:18.01 | d3tul3 | i have almost as many alter egos as you do |
16:18.08 | *** join/#htc-linux ntg (~ntg@141.0.137.20) |
16:19.01 | Cotulla | u didn't fix it so? |
16:20.03 | d3tul3 | Cotulla, you know I can't fix shit in terms of the type of issues you guys are dealing with |
16:20.18 | d3tul3 | it looks like you are making stellar progress though |
16:20.28 | Cotulla | sadly |
16:24.14 | jonpry | i dunno if any progress was made in the last many hours of hacking on it |
16:24.24 | Cotulla | who spelled? |
16:24.32 | Cotulla | looks to everybody |
16:24.38 | jonpry | all i know is when r10 = 0x2f2f2f2f, the shit hits the fan |
16:25.17 | *** join/#htc-linux bartman (~bart@2607:f2c0:f00e:700::dd) |
16:25.22 | jonpry | magic death register |
17:29.15 | d3tul3 | next line something / (r10 - 0x2f2f2f2f), i am sure |
17:29.27 | d3tul3 | i have to tell you everything |
17:29.32 | d3tul3 | where would linux-hexagon be without me |
17:37.05 | jonpry | d3tul3, nowhere |
17:37.08 | jonpry | not in git |
17:51.44 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-85-84.dynamic.mnet-online.de) |
18:23.43 | *** join/#htc-linux ntg (~ntg@141.0.137.20) |
19:05.14 | *** join/#htc-linux helicopter88 (~helicopte@95.234.116.121) |
19:10.40 | *** join/#htc-linux gauner1986 (~Adium@hmbg-5f772a7a.pool.mediaWays.net) |
19:38.40 | *** join/#htc-linux BabelO (~fcr@2a01:e35:8adc:3150:21e:8cff:fe2d:7d3f) |
19:38.43 | *** join/#htc-linux BabelO (~fcr@unaffiliated/babelo) |
20:26.20 | *** join/#htc-linux jonpry (~jon@2602:306:c417:8aa0:d843:2981:bdf1:f8ad) |
23:52.34 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |