Announcement

Collapse
No announcement yet.

X.Org ATI Driver Supports New Power Options

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #46
    sometimes, it doesn't even work /with/ fglrx

    Comment


    • #47
      Originally posted by p4ddY View Post
      How do you read out your GPU's temperature when using one of the open source drivers. aticonfig --od-gettemperature doesn't work without fglrx.
      On my HP nx8220 various status values are exposed via '/proc/acpi/'. So
      Code:
      [val@tux ~]$ cat /proc/acpi/thermal_zone/TZ2/temperature
      temperature:             37 C
      is the temperature of my GPU. TZ1 is the temperature of my CPU.
      If you have a laptop and use powertop on battery you would also see if power saving is in effect when less Watt are consumed.
      On the fact I noted earlier, that without power saving enabled my GPU readily goes up to 60C, I wanted to add that 60C is more or less the temperature my card has when it is under load in fglrx, which somehow suggests that it doesn't idle at all when using the radeon driver without power saving enabled.
      And can I help somehow as tester (but I must say I am non-technical person)? I have a Mobility X600 RV380.

      SignorRossi.

      Comment


      • #48
        Since bugs.freedesktop.org is down, i'm reporting the bug here.

        I have an evil RS482 [Radeon Xpress 200M], and ClockGating and/or DynamicClocks didnt/dont work for me, if i activate it in xorg.conf and restart X, my laptop hangs hard. not even sysrq magic keys can't do a thing. this was true for old dynamicclocks behaviour also
        DynamicPM works but since clockgating is supposed to do the hard work, is suboptimal at this point agd5f feel free to contact me if you want me to debug further this problem.

        best regards and keep up the hard work. btw, does dri2 works for 4xx yet? what do i need? newer drm? do i need mesa with gallium? or the mesa shipped with jaunty works?
        Last edited by groo_pcd; 04-16-2009, 12:12 PM.

        Comment


        • #49
          Originally posted by groo_pcd View Post
          Since bugs.freedesktop.org is down, i'm reporting the bug here.
          Seems to be working fine now.

          Originally posted by groo_pcd View Post
          I have an evil RS482 [Radeon Xpress 200M], and ClockGating and/or DynamicClocks didnt/dont work for me, if i activate it in xorg.conf and restart X, my laptop hangs hard. not even sysrq magic keys can't do a thing. this was true for old dynamicclocks behaviour also
          DynamicPM works but since clockgating is supposed to do the hard work, is suboptimal at this point agd5f feel free to contact me if you want me to debug further this problem.
          I attempted to fix this in git. Let me know how it goes.
          http://cgit.freedesktop.org/xorg/dri...c6bbb08336c83f

          Originally posted by groo_pcd View Post
          best regards and keep up the hard work. btw, does dri2 works for 4xx yet? what do i need? newer drm? do i need mesa with gallium? or the mesa shipped with jaunty works?
          You'll need kms/mm enabled drm, new libdrm, new mesa, new ddx. Easiest method to experiment with it, if you don't want to rebuild everything, is to try Fedora 11.

          Comment


          • #50
            Originally posted by agd5f View Post
            Seems to be working fine now.



            I attempted to fix this in git. Let me know how it goes.
            http://cgit.freedesktop.org/xorg/dri...c6bbb08336c83f
            Ok, its working fine now both dynamic and clockgating thanks a lot

            Originally posted by agd5f View Post
            You'll need kms/mm enabled drm, new libdrm, new mesa, new ddx. Easiest method to experiment with it, if you don't want to rebuild everything, is to try Fedora 11.
            I gather from your answer that dri2 is indeed working with r4xx? Is it usable, aka compiz and xv working? if so, me wanna

            Hmm i know a senior member here in phoronix forums has a PPA with cutting edge X/radeon stuff.. can't remember the name. Maybe he could provide a fresh update to jaunty?

            best regards

            Comment


            • #51
              New bug, since clockgating works now for my rs482, dynamicPM is able to lower my core to 71MHz when i'm on battery mode, BUT when i put the AC back, the clock stays at 71Mhz instead of going up to 400 and plus MHz.

              So, dynamic can lower but can't restore the default state.
              im using rovclock to restore the clock back to 4xxMhz, till a fix emerges.

              best regards.

              UPDATE: https://bugs.freedesktop.org/show_bug.cgi?id=21234
              Last edited by groo_pcd; 04-16-2009, 03:34 PM.

              Comment


              • #52
                Originally posted by groo_pcd View Post

                Hmm i know a senior member here in phoronix forums has a PPA with cutting edge X/radeon stuff.. can't remember the name. Maybe he could provide a fresh update to jaunty?

                best regards
                This one ?
                https://launchpad.net/~tormodvolden/+archive/ppa
                It looks like there are already new packages for Jaunty

                Comment


                • #53
                  Originally posted by Melf View Post
                  This one ?
                  https://launchpad.net/~tormodvolden/+archive/ppa
                  It looks like there are already new packages for Jaunty
                  Thats the one thanks!!! acording to xorgontheedge, the combined ppas + compiled drm should give dri2 to radeon 2xx up to 5xx.

                  anyone did this test already? does it work? i would like to know if it works before trashing my system

                  UPDATE: after i compared the packages from the PPAs i come to the conclusion that jaunty already have more recent packages.
                  so the only thing missing would be drm (which i can compile myself)? is the radeon 4xx dri2 support already in trunk?
                  Last edited by groo_pcd; 04-16-2009, 03:21 PM.

                  Comment


                  • #54
                    Originally posted by groo_pcd View Post
                    New bug, since clockgating works now for my rs482, dynamicPM is able to lower my core to 71MHz when i'm on battery mode, BUT when i put the AC back, the clock stays at 71Mhz instead of going up to 400 and plus MHz.

                    So, dynamic can lower but can't restore the default state.
                    im using rovclock to restore the clock back to 4xxMhz, till a fix emerges.

                    best regards.

                    UPDATE: https://bugs.freedesktop.org/show_bug.cgi?id=21234
                    The DynamicPM option only changes the power state when the screens are blank (DPMS active). If you are getting changes in power states on AC/DC changes, then something else is messing with the chip, perhaps the bios.

                    Comment


                    • #55
                      Originally posted by agd5f View Post
                      The DynamicPM option only changes the power state when the screens are blank (DPMS active). If you are getting changes in power states on AC/DC changes, then something else is messing with the chip, perhaps the bios.
                      I believe you but if its so, the bios wasn't changing anything before i activated the clockgating. anyway, i can confirm the same behaviour when the dpms is activated (back to 71Mhz), doesn't restore the default when the screen comes back.

                      best regards and keep up the good work

                      Comment


                      • #56
                        Originally posted by groo_pcd View Post
                        I believe you but if its so, the bios wasn't changing anything before i activated the clockgating. anyway, i can confirm the same behaviour when the dpms is activated (back to 71Mhz), doesn't restore the default when the screen comes back.

                        best regards and keep up the good work
                        Should be fixed now in git.

                        Comment


                        • #57
                          Originally posted by agd5f View Post
                          Should be fixed now in git.
                          Hum.. :P well.. now it stays at 4xxMhz even with dpms more on.

                          also, now when i go into batery mode it doesnt go down to 71Mhz.

                          Strange indeed. any more fixes?

                          best regards

                          PS. yes i know, my card is evil :P

                          Comment


                          • #58
                            Same issue like groo_pcd, now it's not going down, also for my rV280 it is now slightly overclocked (from 238.5 MHz to 240.75 MHz)

                            Comment


                            • #59
                              It only goes into a low power state when DPMS is active, nothing to do with the AC/DC state. to test:

                              sleep 5; xset dpms force off

                              the chip will drop into a low power state. When you move the mouse, or hit the keyboard, it will switch back up to the higher power state.

                              Comment


                              • #60
                                ok, i can confirm it does work, strange it doesnt change frequency when dpms is called with vbetool, like i was doing. sorry, my mistake.

                                groo@kerberos:~$ sleep 5; xset dpms force off; sudo sh -c 'rovclock -i'
                                Radeon overclock 0.6e by Hasw (hasw@hasw.net)

                                Found ATI card on 01:05, device id: 0x5975
                                I/O base address: 0x9000
                                Video BIOS shadow found @ 0xc0000
                                Reference clock from BIOS: 14.32 MHz
                                Memory size: 131072 kB
                                Memory channels: 0, CD,CH only: 0
                                tRcdRD: 3
                                tRcdWR: 1
                                tRP: 3
                                tRAS: 6
                                tRRD: 1
                                tR2W-CL: 1
                                tWR: 1
                                tW2R: 0
                                tW2Rsb: 0
                                tR2R: 1
                                tRFC: 13
                                tWL(0.5): 0
                                tCAS: 0
                                tCMD: 0
                                tSTR: 0
                                XTAL: 14.32 MHz, RefDiv: 6

                                Core: 71.60 MHz, Mem: 0.0 MHz
                                groo@kerberos:~$ sudo sh -c 'rovclock -i'
                                Radeon overclock 0.6e by Hasw (hasw@hasw.net)

                                Found ATI card on 01:05, device id: 0x5975
                                I/O base address: 0x9000
                                Video BIOS shadow found @ 0xc0000
                                Reference clock from BIOS: 14.32 MHz
                                Memory size: 131072 kB
                                Memory channels: 0, CD,CH only: 0
                                tRcdRD: 3
                                tRcdWR: 1
                                tRP: 3
                                tRAS: 6
                                tRRD: 1
                                tR2W-CL: 1
                                tWR: 1
                                tW2R: 0
                                tW2Rsb: 0
                                tR2R: 1
                                tRFC: 13
                                tWL(0.5): 0
                                tCAS: 0
                                tCMD: 0
                                tSTR: 0
                                XTAL: 14.32 MHz, RefDiv: 6

                                Core: 300.72 MHz, Mem: 0.0 MHz

                                The only thing strange is that, before the last fix, when i was on batery mode, the clock WAS INDEED changing. any explanation on that one? just curious? i know it must be the bios, but it wasnt changing anything before toay rs482 clockgating fix, and i have this laptop for 3 years
                                Last edited by groo_pcd; 04-16-2009, 05:42 PM.

                                Comment

                                Working...
                                X