| 00:18.55 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
| 03:07.06 | *** join/#htc-linux Mavy-bnc (mavfree@nlck1m.ckgaming.co.uk) |
| 03:47.31 | *** join/#htc-linux detule (~detule@pool-96-234-132-77.bltmmd.east.verizon.net) |
| 03:58.52 | *** join/#htc-linux jonpry (~jon@user-0c6tmn0.cable.mindspring.com) |
| 05:23.10 | *** join/#htc-linux {simonsimons34} (~simonsimo@unaffiliated/simonsimons34) |
| 05:32.52 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
| 06:05.28 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
| 06:39.15 | *** join/#htc-linux ALoGeNo (~alogeno@4.Red-79-156-0.staticIP.rima-tde.net) |
| 06:39.16 | *** join/#htc-linux ALoGeNo (~alogeno@unaffiliated/alogeno) |
| 06:58.24 | *** join/#htc-linux ALoGeNo (~alogeno@unaffiliated/alogeno) |
| 08:07.35 | *** join/#htc-linux paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
| 08:35.21 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-84-129.dynamic.mnet-online.de) |
| 08:59.59 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
| 09:01.29 | *** join/#htc-linux eR^zeRa` (~zzeratul@ip-84-42-202-42.net.upcbroadband.cz) |
| 09:17.45 | *** join/#htc-linux arif-ali (~arif-ali@81.144.163.60) |
| 09:39.23 | *** join/#htc-linux arif-ali (~arif-ali@81.144.163.60) |
| 11:01.28 | *** join/#htc-linux BabelO (~fcr@unaffiliated/babelo) |
| 11:28.47 | *** join/#htc-linux helicopter88 (~helicopte@host65-8-dynamic.55-79-r.retail.telecomitalia.it) |
| 11:33.57 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
| 11:50.32 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
| 12:26.04 | *** join/#htc-linux lamikr (lamikr@nat/nokia/x-kgtpxucucfqlouzv) |
| 12:45.17 | *** join/#htc-linux Cotulla (~myfakemai@109.205.253.11) |
| 12:46.52 | Cotulla | hello |
| 12:49.24 | Cotulla | jonpry ping |
| 13:10.34 | jonpry | Cotulla, pong |
| 13:11.02 | Cotulla | it looks like very hard to make relative code for Q6 |
| 13:11.32 | Cotulla | maybe just keep it inside kernel |
| 13:11.41 | Cotulla | but in that case it should not overlap with another TLB entries |
| 13:12.09 | Cotulla | or maybe compile it standalone for fixed address like FFF00000 |
| 13:12.10 | Cotulla | and copy it |
| 13:12.25 | Cotulla | and include as binary into kernel |
| 13:13.04 | Cotulla | also it looks it's rather close to MIPS tlb |
| 13:13.10 | Cotulla | http://cgi.cse.unsw.edu.au/~cs3231/10s1/os161/man/sys161/mips.html |
| 13:15.06 | Cotulla | I honestly like idea with binary more |
| 13:15.37 | Cotulla | it can be included to S file with .incbin directive to kernel, like it's done with zImage |
| 13:17.20 | Cotulla | as well define our interface between binary and kernel in H file |
| 13:17.25 | Cotulla | like at the begining list of values |
| 13:23.06 | jonpry | you want to make the exception handler separate from the kernel? |
| 13:24.30 | jonpry | if the whole kernel is 16mb PTE_G !PTE_U |
| 13:26.20 | jonpry | i've yet to figure out physical address crud |
| 13:26.28 | jonpry | . = PAGE_OFFSET + LOAD_ADDRESS; |
| 13:26.42 | jonpry | afaict the whole kernel is linked at 0xc000 0000 |
| 13:28.13 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
| 13:29.17 | jonpry | hi detule |
| 13:29.26 | detule | hey jonpry |
| 13:29.36 | jonpry | fixed camera? |
| 13:29.53 | detule | yeah the rear camera/video works like a champ |
| 13:30.02 | jonpry | what was it? |
| 13:38.46 | detule | yeah hexagon, i started reading the log from yesterday |
| 13:39.03 | detule | between cotulla's english and your shorthand i think i'll need a bit of time with that log :p |
| 13:39.07 | jonpry | i'm trying to build a qImage |
| 13:39.29 | jonpry | fracking tools don't work with CROSS_COMPILE |
| 13:40.23 | jonpry | or i am doing it wrong |
| 13:40.41 | detule | which toools |
| 13:41.17 | jonpry | https://www.codeaurora.org/patches/quic/hexagon/4.0/ |
| 13:41.39 | jonpry | include/linux/kernel.h:14:20: error: stdarg.h: No such file or directory |
| 13:42.17 | detule | er i don't have happy memories of trying to cross compile nilfs tools |
| 13:42.22 | detule | or anything out-of-kernel |
| 13:42.51 | jonpry | i'm just trying to compile the kernel |
| 13:44.23 | detule | oh these are kernel patches |
| 13:44.51 | detule | chad's releasing his kernel to "donators"
.of course he broke a few things in the process of anti-kanging it |
| 13:45.05 | jonpry | such as? |
| 13:45.33 | detule | don't know that's what his post says
i have a feeling camera might not be working for him |
| 13:47.08 | jonpry | at least the html5 is going |
| 13:47.59 | detule | he is all about these new linaro toolchains
.i guess it's possible that he's not seeing all these header mismatches |
| 13:48.17 | detule | maybe i should upgrade my codesourcery |
| 13:48.34 | jonpry | it's possible your userspace is linaro |
| 13:48.48 | jonpry | i think my stock kernel is linaro compiled |
| 13:48.54 | jonpry | serious size mismatch |
| 13:49.07 | jonpry | gcc = 7.2MB, shipped = 5 |
| 13:49.24 | detule | wait i thought linaro makes them BIGGER |
| 13:49.35 | jonpry | something makes it smaller |
| 13:49.38 | detule | or at least these -O3 flags and what not
. |
| 13:49.51 | detule | i guess that's not a linaro specific flag |
| 13:50.18 | jonpry | then again i have noticed any issues with the 7.2MB kernel |
| 13:50.23 | jonpry | *haven't |
| 13:51.26 | jonpry | Cotulla, how do you compile the kernel with this toolchain? |
| 13:51.43 | jonpry | <PROTECTED> |
| 13:51.46 | jonpry | no good |
| 13:52.29 | Cotulla | sorry I am a bit busy |
| 13:52.39 | Cotulla | reading |
| 13:52.57 | Cotulla | yes, exception handler (tlbmiss) + 6 L1 tables and etc |
| 13:53.07 | Cotulla | 16M can be too big waste of physical memory |
| 13:53.39 | Cotulla | detule, my second nick is englishpoo :P |
| 13:54.07 | Cotulla | jonpry, I copied boot folder from ARM |
| 13:54.12 | Cotulla | to make qImage from ELF |
| 13:54.22 | Cotulla | CROSS_COMPILE works for me for kernel at Linux32 |
| 13:54.26 | Cotulla | oh |
| 13:54.31 | Cotulla | stdarg.h |
| 13:54.34 | jonpry | yeah |
| 13:54.48 | Cotulla | on CA it's a bit fake release |
| 13:54.52 | Cotulla | they excluded all headers actually |
| 13:54.56 | Cotulla | witches |
| 13:55.04 | jonpry | ugh |
| 13:55.09 | Cotulla | but u need only stdarg.h for kernel |
| 13:55.13 | Cotulla | I copied it from somewhere |
| 13:55.26 | Cotulla | or maybe a bit more H files, but almost nothing else is used |
| 13:56.01 | detule | lol fake release |
| 13:56.02 | Cotulla | yes I did something like "ARCH=hexagon CROSS_COMPILE=/opt/gnu/bin/hexagon-" |
| 13:56.14 | Cotulla | but I put that inside Makefile as usual |
| 13:56.30 | Cotulla | do u get other errors except stdarg.h? |
| 13:56.49 | jonpry | not yet |
| 14:00.10 | jonpry | apparently the nexus4 kernel does not build cleanly for hexagon |
| 14:08.52 | jonpry | Cotulla, are you building kernel.org 3.4? |
| 14:09.12 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
| 14:26.07 | Cotulla | hm |
| 14:26.11 | Cotulla | I think yes |
| 14:26.19 | Cotulla | how to check GIT source path? |
| 14:26.55 | jonpry | .git/config ? |
| 14:37.41 | Cotulla | ok |
| 14:38.51 | {simonsimons34} | hum. okay so i need to port mddi drivers to lcdc for my device. any helpful guidance |
| 14:45.16 | Cotulla | jonpry, u want to know or not really?> |
| 14:45.20 | Cotulla | actually it's old git |
| 14:45.26 | Cotulla | I think they made some changes inside already |
| 14:45.34 | Cotulla | but looks like in 3.8 it is not really changed |
| 14:50.45 | Cotulla | simonsimons34 which device? |
| 14:51.04 | jonpry | maybe we should just set something up |
| 14:51.21 | jonpry | since i am ready to start building kernels |
| 14:51.32 | jonpry | and you have some codez |
| 14:51.38 | Cotulla | not yet working :) |
| 14:52.14 | jonpry | thats fine |
| 14:52.32 | jonpry | i have to get it to compile for v4 ELF |
| 14:52.35 | Cotulla | assembler is a different from ARM |
| 14:52.54 | Cotulla | so we didn't finish discussion about |
| 14:53.06 | Cotulla | "kernel manager" |
| 14:53.10 | Cotulla | or how to call it |
| 14:53.57 | jonpry | you mean like a fake hypervisor? |
| 14:54.03 | Cotulla | kinda yes |
| 14:54.07 | Cotulla | but it's not really hypervisor :) |
| 14:54.15 | jonpry | i thought we were going to implement the stuff in the kernel |
| 14:54.18 | {simonsimons34} | Cotulla, im working on the one v |
| 14:54.20 | Cotulla | or SW MMU |
| 14:54.36 | Cotulla | well |
| 14:54.40 | Cotulla | so maybe map kernel as 1M pages |
| 14:54.43 | Cotulla | and one page will be fixed |
| 14:54.47 | jonpry | you know how kexec stuff works right? |
| 14:55.00 | Cotulla | hm heard |
| 14:55.01 | jonpry | there is an ELF section for special purpose |
| 14:55.09 | Cotulla | hm |
| 14:55.11 | jonpry | idmap_section iirc |
| 14:55.16 | Cotulla | oh |
| 14:55.26 | Cotulla | but it's important to keep that section outside of main TLB |
| 14:55.36 | Cotulla | I don't think it will like collision |
| 14:55.52 | jonpry | maybe |
| 14:55.54 | Cotulla | and it's not possible to copy code without processing relocations |
| 14:56.24 | jonpry | i don't think it needs relocation |
| 14:56.27 | Cotulla | or maybe just specify relocations by hands |
| 14:56.35 | jonpry | just linked correctly in the beginning |
| 14:56.44 | Cotulla | hm |
| 14:56.45 | jonpry | add an elf section to link at 0xc100 0000 |
| 14:56.52 | jonpry | or whatever |
| 14:56.59 | jonpry | virtual memory is cheap |
| 14:57.01 | Cotulla | ever at 0xFFF00000? |
| 14:57.08 | jonpry | sure |
| 14:57.08 | Cotulla | in that case it's good |
| 14:57.16 | Cotulla | hm add it to LDS ? |
| 14:57.21 | jonpry | yeah |
| 14:57.34 | Cotulla | well |
| 14:57.35 | jonpry | just like they did for kexec trampoline |
| 14:57.37 | Cotulla | sounds good |
| 15:06.11 | jonpry | /opt/gnu/bin/hexagon-ld: libgcc.a: No such file: No such file or directory |
| 15:06.12 | jonpry | make: *** [vmlinux.o] Error 1 |
| 15:10.16 | Cotulla | hm |
| 15:10.19 | Cotulla | exclude it? |
| 15:10.22 | Cotulla | I did it so I think |
| 15:10.30 | Cotulla | but I added helpers by hands |
| 15:10.58 | Cotulla | division and etc |
| 15:12.44 | *** join/#htc-linux helicopter88_2 (~helicopte@host4-116-dynamic.55-79-r.retail.telecomitalia.it) |
| 15:13.02 | jonpry | i don't see where it is included for hexagon |
| 15:15.24 | *** join/#htc-linux helicopter88_2 (~helicopte@host212-82-dynamic.37-79-r.retail.telecomitalia.it) |
| 15:16.24 | Cotulla | hm |
| 15:16.30 | Cotulla | it can be standard lib |
| 15:16.38 | Cotulla | just put some shitty lib as libgcc maybe |
| 15:16.44 | Cotulla | I don't remember my solution |
| 15:22.07 | *** join/#htc-linux helicopter88 (~helicopte@host167-115-dynamic.23-79-r.retail.telecomitalia.it) |
| 15:31.20 | *** part/#htc-linux Cotulla (~myfakemai@109.205.253.11) |
| 15:31.25 | *** join/#htc-linux Cotulla (~myfakemai@109.205.253.11) |
| 15:31.29 | Cotulla | lol back |
| 15:32.43 | Cotulla | so did u fix it jonpry? |
| 15:32.45 | Cotulla | ;) |
| 15:33.17 | Cotulla | I have that kernel on another PC and I need to boot it to look into. so I can't do that fast. |
| 15:38.05 | jonpry | i got it to not try to link |
| 15:38.14 | jonpry | but all kinds of undefined refs then |
| 15:38.56 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
| 15:39.43 | *** join/#htc-linux a-st (~a-st@unaffiliated/a-st) |
| 15:45.14 | jonpry | i found the source for it |
| 15:46.11 | jonpry | lib1funcs.asm |
| 15:46.41 | Cotulla | yea |
| 15:46.44 | Cotulla | I copied it too |
| 15:47.11 | jonpry | this thing sucks |
| 15:47.24 | jonpry | compile time hexagon arch |
| 15:47.40 | Cotulla | it should be fixed once |
| 15:47.43 | Cotulla | and done |
| 15:47.52 | Cotulla | actually I prefer to have .S file in kernel |
| 15:47.53 | jonpry | can it be added to the kernel source |
| 15:47.56 | jonpry | ? |
| 15:47.59 | Cotulla | yes I think |
| 15:48.06 | Cotulla | instead of some external lib |
| 15:48.11 | jonpry | yeah |
| 15:48.15 | jonpry | that is arch specific |
| 15:48.25 | Cotulla | yes |
| 15:48.27 | Cotulla | division |
| 15:48.30 | Cotulla | 64 bits |
| 15:48.33 | Cotulla | and helpers for another things |
| 15:48.45 | Cotulla | like ARM once |
| 15:52.10 | Cotulla | in my case |
| 15:52.12 | Cotulla | I also changed |
| 15:52.16 | Cotulla | QDSP6 <-> HEXAGON |
| 15:52.17 | Cotulla | names |
| 15:54.55 | jonpry | i'd prefer not to reinvent the wheel |
| 15:55.12 | jonpry | are you going to post this stuff relatively soon? |
| 15:56.06 | Cotulla | which stuff? |
| 15:56.16 | Cotulla | I think this kernel is outdated |
| 15:56.32 | Cotulla | problem that they renamed compiler |
| 15:56.36 | Cotulla | from qdsp6 to hexagon |
| 15:56.51 | Cotulla | so some compilers are using qdsp6* helpers and another more modern hexagon* |
| 15:59.53 | jonpry | should i just pull a 3.7 kernel? |
| 16:03.44 | Cotulla | not 3.8? |
| 16:04.48 | jonpry | rc kernels suck |
| 16:06.55 | Cotulla | ok |
| 16:07.08 | Cotulla | has no idea what is lastest linux version |
| 16:07.23 | jonpry | 3.7.6 or 3.8-rc7 |
| 16:07.28 | Cotulla | oh |
| 16:07.33 | Cotulla | yes maybe 3.7 |
| 16:07.38 | Cotulla | can we run android on it later? |
| 16:07.51 | jonpry | probably |
| 16:12.37 | jonpry | detule, are you taking charge of the git? |
| 16:14.50 | *** join/#htc-linux eR^zeRa` (~zzeratul@ip-84-42-202-42.net.upcbroadband.cz) |
| 16:30.03 | Cotulla | I will continue soon developing of tlb miss handler |
| 16:30.16 | Cotulla | I stopped on conversion from HVM entry to TLB format |
| 16:30.27 | Cotulla | need translate all that sizes to mask |
| 16:32.16 | detule | jonpry, i can create that repo 3.7 stable on github? it may have to wait til monday though |
| 16:41.14 | Cotulla | hm |
| 16:41.17 | Cotulla | for q6? |
| 16:43.24 | jonpry | Cotulla, yes hexagon |
| 16:43.44 | jonpry | detule, yes 3.7 stable |
| 16:43.45 | Cotulla | good idea to create git |
| 16:44.03 | Cotulla | is github fast enough? |
| 16:44.32 | jonpry | seems pretty good. anything but the initial push/pull is very fast anyways |
| 16:44.45 | Marc | it's definitely faster than gitorious ;) |
| 16:45.15 | Cotulla | ok |
| 16:50.09 | Cotulla | btw jonpry u talked about 7 L1, why not 6? |
| 16:50.39 | Cotulla | I think everyone should has kernel mappings too, to make code more easy and fast |
| 16:50.57 | jonpry | kernel should just be a separate table |
| 16:51.05 | Cotulla | why? |
| 16:51.12 | Cotulla | it's already handled inside linux kernel code |
| 16:51.19 | jonpry | otherwise threads will be evicting kernel TLB entries from other threads |
| 16:51.22 | Cotulla | better to have full 1024 entries L1 |
| 16:52.04 | Cotulla | it's easy to index inside such table |
| 16:52.08 | Cotulla | 10 bits from address to L1 |
| 16:52.15 | Cotulla | another 10 bits from address to L2 |
| 16:52.32 | Cotulla | but we are going to use ASID |
| 16:52.41 | Cotulla | to make different mappings for each thread |
| 16:52.51 | Cotulla | or maybe mark kernel ones as global |
| 16:52.57 | Cotulla | and make them standalone |
| 16:53.26 | jonpry | yes i think global for kernel |
| 16:53.37 | Cotulla | maybe 1 + 3 + 6 * 8 + 12 |
| 16:53.39 | Cotulla | 12 for kernel |
| 16:54.03 | Cotulla | or 6 * 9 and 6 for kernel |
| 16:54.22 | jonpry | i think 6*8 should just be 48 FILO |
| 16:54.32 | Cotulla | without care about thread id? |
| 16:54.36 | jonpry | yeah |
| 16:54.45 | Cotulla | yes also possible variant |
| 16:54.47 | jonpry | so single thread can take over the whole TLB |
| 16:55.03 | Cotulla | hm yes |
| 16:55.21 | Cotulla | so 48 or 54 and 12 / 6 for kernel |
| 16:55.36 | jonpry | yes |
| 16:55.39 | Cotulla | or maybe make them also together |
| 16:55.42 | Cotulla | just 60 entries |
| 16:56.17 | Cotulla | then 1 TLB for "system core" (64K) |
| 16:56.26 | Cotulla | and 3 TLB for L2 mapping |
| 16:56.48 | Cotulla | or maybe just one |
| 16:57.02 | Cotulla | or 6 |
| 16:57.05 | Cotulla | one per thread |
| 16:57.13 | Cotulla | I tested already |
| 16:57.24 | Cotulla | it's not possible to generate another tlbmiss from tlbmiss code |
| 16:57.28 | Cotulla | u got error exception |
| 16:57.47 | Cotulla | so tlbmiss should always map L2 before access |
| 16:58.14 | Cotulla | there is also tlbp instruction |
| 16:58.20 | Cotulla | which checks mapping |
| 16:59.11 | Cotulla | but it works with VA address |
| 16:59.29 | Cotulla | or maybe store in global variable |
| 16:59.32 | Cotulla | last mapping |
| 16:59.36 | Cotulla | and compare |
| 16:59.44 | Cotulla | to prevent setting same mapping |
| 16:59.52 | jonpry | some conditions allow nested tlbmiss |
| 17:00.03 | jonpry | qcom patent uses that technique |
| 17:00.19 | Cotulla | hm it's not working for me during tests |
| 17:00.40 | jonpry | say all kernel l1 and l2 is in TLB |
| 17:00.49 | jonpry | then a miss on reading userspace l2 |
| 17:01.03 | jonpry | which is located in kernel memory |
| 17:01.08 | Cotulla | but maybe |
| 17:01.14 | jonpry | should work. assuming you have implemented a stack |
| 17:01.16 | Cotulla | it's possible to do like in mips |
| 17:01.21 | *** join/#htc-linux arif-ali (~arif-ali@94-192-24-56.zone6.bethere.co.uk) |
| 17:01.27 | Cotulla | handle tlbmiss inside error handler |
| 17:02.12 | Cotulla | but it's not working . . . I tried to access invalid memory |
| 17:02.14 | Cotulla | inside tlbmiss |
| 17:02.17 | Cotulla | and got error exception |
| 17:03.01 | jonpry | did you reenable exceptions? |
| 17:03.07 | Cotulla | no |
| 17:03.25 | jonpry | i think that is required |
| 17:03.34 | Cotulla | how? |
| 17:03.34 | Cotulla | :D |
| 17:04.01 | Cotulla | I am afraid it can be not possible |
| 17:04.05 | Cotulla | I don't have examples |
| 17:04.25 | jonpry | accessing memory that could result in a tlbmiss requires that exceptions be enabled |
| 17:05.00 | jonpry | probably SSR bit |
| 17:05.58 | Cotulla | well it can assume that no tlbmiss exceptions should be inside tlbmiss handler |
| 17:06.01 | Cotulla | looks logical |
| 17:06.31 | jonpry | unless you want nested exceptions |
| 17:06.52 | Cotulla | yes but they can be unsupported |
| 17:07.50 | jonpry | it is supported |
| 17:07.59 | Cotulla | I am not sure |
| 17:08.10 | jonpry | did you read the patent? |
| 17:08.15 | Cotulla | no |
| 17:08.37 | jonpry | http://www.google.pl/patents/download/7398371_Shared_translation_look_aside_bu.pdf?id=_mirAAAAEBAJ&output=pdf&sig=ACfU3U3IfaK90rO1CsP0UU3v7PFUm4EBEg&source=gbs_overview_r&cad=0 |
| 17:09.33 | Cotulla | lol |
| 17:09.35 | Cotulla | funny one |
| 17:09.40 | jonpry | yeah |
| 17:09.49 | Cotulla | but I have no idea how to implement it in code |
| 17:10.06 | jonpry | worst case just enl = pc+1, rte |
| 17:10.14 | jonpry | er eln = pc+1, rte |
| 17:10.22 | Cotulla | elr |
| 17:10.29 | jonpry | :p |
| 17:10.46 | Cotulla | well |
| 17:10.55 | Cotulla | I think better start from just mapping L2 |
| 17:10.58 | Cotulla | as TLB entry |
| 17:11.30 | Cotulla | well few tlbmissed can make a dead loop is not it? |
| 17:11.38 | Cotulla | *tlbmisses |
| 17:11.52 | jonpry | must be careful |
| 17:12.15 | jonpry | i think it works because a threads TLB is located in the permanently mapped kernel l1/l2 |
| 17:12.27 | jonpry | so worst case is a single nested fault |
| 17:13.26 | Cotulla | well |
| 17:13.35 | Cotulla | ARM always using physical addresses to access entries |
| 17:13.44 | Cotulla | while in our case we need virtual mapping |
| 17:13.53 | jonpry | yes |
| 17:13.56 | Cotulla | ever ARM can work if L2 entry is not mapped to VA |
| 17:14.15 | Cotulla | and looks like HVM PTE stores PA address |
| 17:14.27 | Cotulla | so we need map it anyway to take value |
| 17:14.47 | Cotulla | we have no idea where is it mapped to VA however |
| 17:15.50 | jonpry | i'm a little lost |
| 17:16.12 | Cotulla | well |
| 17:16.23 | Cotulla | HVM L1 and L2 they contain PA addresses right? |
| 17:16.39 | Cotulla | so at TLBMISS |
| 17:16.49 | Cotulla | or not |
| 17:16.50 | jonpry | the entries are like both VA and PA no? |
| 17:16.51 | Cotulla | hm |
| 17:16.53 | Cotulla | I am lost too :D |
| 17:17.03 | Cotulla | look at document |
| 17:17.06 | Cotulla | HVM specifications |
| 17:17.19 | Cotulla | 9 MemMgr |
| 17:17.27 | Cotulla | 9.4 |
| 17:17.44 | Cotulla | hm |
| 17:17.48 | Cotulla | page 35 |
| 17:17.58 | Cotulla | "logical address of L2 table" |
| 17:18.07 | Cotulla | is it VA or PA? |
| 17:18.09 | jonpry | i need to look at this a little later. i'm trying to fix an electrical problem in my car |
| 17:18.17 | Cotulla | well |
| 17:18.20 | Cotulla | 9.4.2 |
| 17:18.37 | Cotulla | "PTEs are used to define the virtual-to-logical translaton . . " |
| 17:18.44 | Cotulla | so logical it's like PA |
| 17:18.51 | Cotulla | and L1 has PA of L2 |
| 17:19.02 | Cotulla | so we need map it anyway |
| 17:19.04 | Cotulla | to get access |
| 17:19.10 | Cotulla | map to some VA |
| 17:20.16 | Cotulla | ok |
| 17:20.21 | Cotulla | I gotta away for some time |
| 17:20.23 | Cotulla | cya |
| 17:20.45 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
| 17:21.06 | detule | jonpry, you care about seeing the prior commit history of 3.7 or just init the repo from the tarball |
| 17:22.52 | Cotulla | better without histroy |
| 17:26.40 | detule | Cotulla, you have a github account |
| 17:26.42 | detule | ? |
| 17:29.07 | Cotulla | not yet or maybe had but forgot |
| 17:29.15 | Cotulla | I will make a new one |
| 17:29.19 | Cotulla | now real away |
| 17:34.43 | jonpry | go tarball |
| 17:40.53 | detule | you should have it |
| 17:52.06 | jonpry | awesome, thanks |
| 18:01.30 | *** join/#htc-linux ElFinLazz (~elfinlazz@182.215.84.22) |
| 18:24.01 | Cotulla | back |
| 18:24.17 | Cotulla | how is car jonpry? |
| 18:35.13 | Cotulla | detule |
| 18:35.16 | Cotulla | how is it? |
| 18:52.12 | detule | Cotulla, got that account |
| 18:53.29 | Cotulla | ? |
| 18:53.59 | *** join/#htc-linux rajkosto (~rajkosto@wan.rajkonet.info) |
| 18:54.19 | detule | github did you make it |
| 18:55.14 | Cotulla | oh |
| 20:07.19 | Cotulla | hm I assume L2 must be always 4096 bytes aligned right? |
| 20:30.34 | *** join/#htc-linux stroughtonsmith (~steven@86-42-150-7-dynamic.b-ras1.bbh.dublin.eircom.net) |
| 20:38.26 | *** join/#htc-linux Rajko (~rajkosto@wan.rajkonet.info) |
| 20:42.01 | *** join/#htc-linux wk (~123123@141.0.137.20) |
| 23:12.14 | *** join/#htc-linux khorben__ (~dont@91.102.9.98) |
| 23:24.13 | *** join/#htc-linux detule (~detule@pool-96-234-132-77.bltmmd.east.verizon.net) |
| 23:46.50 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |