#tyndur · 2010-02-10

779 Nachrichten · Seite 2 von 4 · Original-CSV herunterladen

  1. eddyband no luck with waiting, btw
  2. XanClicDommage
  3. eddybback
  4. eddybis-a-geek.org?
  5. eddybah, dyndns
  6. eddybi like that one :)
  7. eddybbtw, you know what i did to the usb-hcd module?
  8. XanClicNo, not yet
  9. eddybmaybe the expression doesn't work in english... you should've asked "what?" :P
  10. eddybhttp://codepad.org/CHHyRUNp << enjoy
  11. XanClicHm, I'm sorry, but I don't know the old version ;-)
  12. XanClic"searchByClassSubclassAndProgInterface"
  13. XanClicI like those names xD
  14. eddybnow i see i forgot to do g_GotXhci = true; in the Xhci handler... is not implemented, anyway
  15. eddybheh
  16. eddybi did those :P
  17. XanClicOverloading is possible in C++, isn't it?
  18. eddybi had been told i can use whatever i like
  19. eddybwhat overloading?
  20. XanClicWhat about a function called "search_by_class"?
  21. eddybyuck
  22. XanClicIf you give one parameter, you search by class
  23. XanClicTwo parameters -> class and subclass
  24. XanClicThree -> class, subclass, programming interface
  25. eddyblowerCamelCase :)
  26. eddybthe callback function after that?
  27. xan_clicCamelCase is bad ;-)
  28. eddybwe use it everywhere
  29. eddybi wanted it in the pedigree directory structure(/Applications is more modern than /applications :P), but matt said no... but i think it'd be ok when we get a GUI
  30. kwolfCapital letters are harder to type
  31. XanClicI'll call my GUI g_u_i *g*
  32. eddyband it should be obvious "what's up"... anyway, the old one is http://www.pedigree-project.org/projects/pedigree/repository/revisions/master/entry/src/modules/drivers/common/uhci/main.cc
  33. eddybI added xHCI!!!
  34. XanClicAh ;-)
  35. XanClicHm, C++ is not a nice thing to look at
  36. XanClicIt looks like C, but it isn't
  37. XanClicI like Sather, because it looks completely different :D
  38. kwolfYou should write your next kernel in Brainfuck
  39. XanClicThat's not object-oriented ;-)
  40. XanClicC also looks like C, but I like it ;-)
  41. kwolfYou can write nice code as well as ugly code in almost any language
  42. kwolfThough there are languages which make the nice part hard
  43. XanClicI just don't like reading a language that pretends to be another one which it isn't *g*
  44. XanClic(Hm, I think I'm lucky bluecode isn't here and it's 23:56 at pcmattman)
  45. kwolfThey don't have op. Rather you're lucky that tyndur isn't written in C++. ;-)
  46. XanClic*g*
  47. kwolf(tyndur, the bot, not the OS)
  48. XanClicAh ;-)
  49. kwolfYay, I think I have achieved my goal!
  50. kwolf"42 errors were found on the image."
  51. XanClicAnybody with a nice VFS concept for a monolithic kernel out there? Thou shalt speak now or stay quiet forever ;-)
  52. XanClicWell, I think my old English isn't the yellow from the egg
  53. XanClickwolf: lol
  54. kwolfXanClic: You mean outlanders understand only railway station? ;-)
  55. kwolfThe nice VFS concept is called LIOv2, of course *g*
  56. XanClicExactly that I mean ;-) - time for some language rape
  57. XanClickwolf: stress on _monolithic_ ;-)
  58. kwolfI'm perfectly able to read ;-)
  59. XanClic*g*
  60. kwolfSo, yes, LIOv2
  61. XanClicOuuh, too late ;-)
  62. kwolfUsing it is going to be much easier in a monolithic kernel
  63. kwolfFor what reason, my dear XanClic, dost thou ask when thou art not even interested in the answer? ;-) (hm, better word for "interested", anyone?)
  64. XanClics/the/that/ ;-)
  65. kwolfThat's not the word I asked thee to replace ;-)
  66. XanClicThou shalt obey the second part of my "proposal" ;-)
  67. XanClicThough not forever *g*
  68. kwolfPah. ;-)
  69. kwolfYay, KDE 2 window decorations are fixed! :-)
  70. XanClicxD
  71. kwolfFinally my windows look again like they should
  72. eddybkde2?
  73. eddyb=))
  74. kwolfYup, don't like most of the newer styles
  75. eddybas i said earlier, you are making progress harder :P
  76. kwolfThough Quartz didn't look too bad as an ersatz style while the KDE 2 one was broken
  77. XanClic"ersatz style" is nice
  78. kwolfIf I understand correctly, it means exactly what I wanted to say ;-)
  79. XanClicI didn't mark it as wrong ;-)
  80. kwolfI guess we'll need a native speaker to review the log of today ;-)
  81. kwolfThough finding a native speaker for thy old languange might prove difficult ;-)
  82. XanClicJanDing invented a time machine yesterday, didn't he? ;-)
  83. eddyb?
  84. XanClicSomebody in #lost described yesterday how to finish a program before it has started
  85. eddyb??
  86. kwolfXanClic: That were JanDing ;-)
  87. XanClic*g*
  88. eddybBetriebssysteme << what's this?
  89. XanClicOperating systems
  90. kwolfXanClic: That's easy actually. Run bochs in valgrind. It's so slow that it's going to run backwards.
  91. XanClicxD
  92. XanClicI refuse to run bochs ;-)
  93. kwolfFunny. Now eddyb joined #lost before quok did. Even though he knows less German, I suspect.
  94. XanClic*g*
  95. XanCliceddyb: He said, "For i=1 to 20" (qbasic) would be much more beautiful than "for (i = 0; i < 10; i++)", furthermore it does 20 loops instead of only 10
  96. XanClicI then told him that it's also twice as slow
  97. XanClicSo he said, looping backwards ("step -1" or "i--") would result in a for loop that lasts negative time
  98. XanClicThus the program finishs before it starts
  99. eddybwhatever
  100. eddybhttp://forum.osdev.org/viewtopic.php?f=11&t=21531&p=172155#p172128 << Bryten Inc and his OpenBSD XD
  101. eddybwell, i couldn't rezist replying :P
  102. XanClicopenBSD is cool, they have nice release songs :D
  103. eddybhehe
  104. eddybthen i have to add support for sound in the CDI version...
  105. eddybduh, if there was a BIG bunch of drivers, i would like CDI...
  106. eddybi don't know if there's a driver we use *and* is functional...
  107. XanClicThe network drivers are functional
  108. eddybyeah, right
  109. eddybe1000 halts
  110. kwolfAnd the ac97 storage driver works, too, right? ;-)
  111. XanClicBut it's not official yet ;-)
  112. kwolfI might just commit it ;-)
  113. XanClicIt works, yes ;-)
  114. kwolfIn qemu at least
  115. kwolfNever tried it on real hardware
  116. XanClicAh, damn, I knew I forgot something xD
  117. eddybargh, i should've done that before you...
  118. eddybnow i have to use CDI for it
  119. eddybXD
  120. kwolfAt some point I'll provide patches to replace all your native drivers ;-)
  121. eddybyou know by then you're gonna be shot
  122. XanClicHm, I don't know yet if I'm going to implement CDI in paloxena3...
  123. eddybkwolf: ah, and i think matt won't let you, it's regression
  124. kwolfeddyb: By whom? ;-)
  125. eddybby me....
  126. eddybif i have to come in germany to do that
  127. XanClicBy the way, where are you from?
  128. eddybRomania
  129. kwolfeddyb: Meh, so finally Schengen will prove to be a mistake
  130. XanClicAh
  131. XanClicxD
  132. kwolfNot sure why it should be a regression, though
  133. kwolfHm 2 errors of the 42 left. I don't like that number, I should revert.
  134. eddyblol?
  135. eddybi just made it print the frame number in a loop, delayed by a few milisec...
  136. eddybit continues to parse the frame list...
  137. eddyband IOC+active are unset
  138. quokmoin
  139. quokgot a nice backlog to read I see!
  140. eddybmoin quok
  141. XanClicMoin, quok
  142. kwolfMoin quok
  143. kwolfDon't think it's really worth reading ;-)
  144. eddybme neither
  145. eddybhat reminds me to check the #pedigree one
  146. eddyb*that
  147. kwolfIs there even a backlog?
  148. eddybhttp://irclogs.quokforge.org/pedigree/ ?
  149. eddybor you mean no one talked?
  150. kwolfThat's what I meant, yes
  151. eddybah
  152. eddybwell, only matt saying good night everybody :P
  153. kwolfBetter than nothing ;-)
  154. eddybah, funny, wrong ordering of the values in NOTICE
  155. eddyblike doesn't even touch the entry...
  156. eddybargh
  157. eddybi got it...
  158. eddybwell, half of it..
  159. eddybis not triggering an interrupt because it doesn't need to...
  160. eddybaka it reads from wrong address or something
  161. quokfor a minute I thought I was in #pedigree, but this is #tyndur...
  162. eddyblol
  163. kwolfYou should have known. This is the channel with some activity. ;-)
  164. eddybthis is because i was talking with XanClic about what could cause my UHCI problem
  165. eddybheh
  166. eddybsome activity *besides eddyb*
  167. kwolfYay, rewrote half of the refcount handling and the test suite passes again :-)
  168. eddybtest suite?
  169. kwolfRefcount handling of qcow2 and the test suite is qemu-iotests (I doubt that you know it)
  170. eddybyou're working on qemu, then?
  171. kwolfYes
  172. eddybXanClic: suppose we got a nasty controller, like VMware's... what should it complain about most?
  173. eddybit either doesn't read from right address, either it doesn't like the frame list...
  174. XanCliccomplain?
  175. XanClicHm, don't know
  176. eddybWTF?
  177. eddybi managed to crash VMware....
  178. eddybah
  179. eddybwhoops
  180. eddybthe write[8|16|32](value, offset) syntax is a bit confusing some times
  181. kwolfPedigree API?
  182. eddybyup
  183. eddybm_pBase->write16(UHCI_CMD_RUN, UHCI_CMD);
  184. kwolfHm, I'm used to address, value
  185. eddybit'd be easier for me when i do quick changes with copy&paste...
  186. eddyb(NN) 0x1800000 0x0 0x1 0x58c000 0x0 << looks fine, the values are "pTD->ctrl(active+IOC) STS(nothing) CMD(run) FRAME_LIST_POINTER(looks good, i ussually get this value for the first allocation) FRAME_NUMBER(0, is the first in the loop)"
  187. eddyband it doesn't touch the TD XD
  188. eddybfunny, why i'm doing m_pFrameList[1]=m_pTDListPhys; ?? it should be index 0
  189. eddybWTFWTF?
  190. eddybi finally get an IRQ
  191. eddyba baad irq
  192. kwolfThere are good and bad IRQs?
  193. eddyblike Host Controller Process Error is set :P
  194. quokOoooh, bewing sent me a tarball of his bochs rewrite
  195. eddybokey...
  196. eddybin Uhci::setup i've got interrupts disabled....
  197. eddybhow come that?
  198. eddybwhere's matt when i need him XD
  199. XanClicIn bed ;-)
  200. eddybexactly my point...