Re: drm2 regression: backlight adjustment on ivybridge no longer works
On Sun, Feb 01, 2015 at 11:23:47AM +0100, Lars Engels wrote: On 1. Februar 2015 11:20:26 MEZ, Alexandr Krivulya shur...@shurik.kiev.ua wrote: 01.02.2015 12:12, Lars Engels пишет: On Sun, Feb 01, 2015 at 10:41:25AM +0100, Lars Engels wrote: On Sun, Feb 01, 2015 at 03:35:15AM -0600, Andrew Wilcox wrote: Lars Engels sent: 01 February 2015 03:18: With acpi_video I get some interesting sysctl: hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 I guess it should not be 100 100 0 ... 100? Actually, the standard internal ACPI brightness level struct (BRTN in the DSDT) is laid out as: * Full power value (one byte, the value at which the brightness should be set on AC by default) * Economy value (one byte, the value at which the brightness should be set on battery by default) * Actual values (N bytes, up to Max but frequently not) So, no, that value indeed sounds correct. On my laptop the value is: 80 47 0 7 13 20 27 33 40 47 53 60 67 73 80 87 93 100 Thank you for the explanation. FWIW, here is the full output from sysctl hw.acpi.video.lcd0: hw.acpi.video.lcd0.active: 1 hw.acpi.video.lcd0.brightness: 11 hw.acpi.video.lcd0.fullpower: 100 hw.acpi.video.lcd0.economy: 100 hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 hw.acpi.video.crt0.active: 1 hw.acpi.video.ext0.active: 1 hw.acpi.video.ext1.active: 1 hw.acpi.video.ext2.active: 1 hw.acpi.video.ext3.active: 1 hw.acpi.video.ext4.active: 1 hw.acpi.video.ext5.active: 1 What revision of -CURRENT are you running? What is the outcome of trying the patch posted Saturday morning (UTC) from Elizabeth Myers (message ID 54cc5311.9070...@interlinked.me)? I'm currently running r277858 and haven't tried the patch, yet. But will do now and report back. So now I have a new sysctl hw.dri.0.i915_backlight. But no matter what value I pass to it, the brightness doesn't change. It's a Ivy Brigde notebook where drm shows some errors in dmesg. Please see: http://pastie.org/9877917 Please try to boot with acpi_ibm and without acpi_video. This way all works for me. That‘s already what I do. Also tried it with acpi_video without luck. For the records: With r278102 brightness controls are working again. Thanks all for helping. pgpRxf8xrTJKt.pgp Description: PGP signature
Re: drm2 regression: backlight adjustment on ivybridge no longer works
01.02.2015 12:12, Lars Engels пишет: On Sun, Feb 01, 2015 at 10:41:25AM +0100, Lars Engels wrote: On Sun, Feb 01, 2015 at 03:35:15AM -0600, Andrew Wilcox wrote: Lars Engels sent: 01 February 2015 03:18: With acpi_video I get some interesting sysctl: hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 I guess it should not be 100 100 0 ... 100? Actually, the standard internal ACPI brightness level struct (BRTN in the DSDT) is laid out as: * Full power value (one byte, the value at which the brightness should be set on AC by default) * Economy value (one byte, the value at which the brightness should be set on battery by default) * Actual values (N bytes, up to Max but frequently not) So, no, that value indeed sounds correct. On my laptop the value is: 80 47 0 7 13 20 27 33 40 47 53 60 67 73 80 87 93 100 Thank you for the explanation. FWIW, here is the full output from sysctl hw.acpi.video.lcd0: hw.acpi.video.lcd0.active: 1 hw.acpi.video.lcd0.brightness: 11 hw.acpi.video.lcd0.fullpower: 100 hw.acpi.video.lcd0.economy: 100 hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 hw.acpi.video.crt0.active: 1 hw.acpi.video.ext0.active: 1 hw.acpi.video.ext1.active: 1 hw.acpi.video.ext2.active: 1 hw.acpi.video.ext3.active: 1 hw.acpi.video.ext4.active: 1 hw.acpi.video.ext5.active: 1 What revision of -CURRENT are you running? What is the outcome of trying the patch posted Saturday morning (UTC) from Elizabeth Myers (message ID 54cc5311.9070...@interlinked.me)? I'm currently running r277858 and haven't tried the patch, yet. But will do now and report back. So now I have a new sysctl hw.dri.0.i915_backlight. But no matter what value I pass to it, the brightness doesn't change. It's a Ivy Brigde notebook where drm shows some errors in dmesg. Please see: http://pastie.org/9877917 Please try to boot with acpi_ibm and without acpi_video. This way all works for me. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
On Sun, Feb 01, 2015 at 10:41:25AM +0100, Lars Engels wrote: On Sun, Feb 01, 2015 at 03:35:15AM -0600, Andrew Wilcox wrote: Lars Engels sent: 01 February 2015 03:18: With acpi_video I get some interesting sysctl: hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 I guess it should not be 100 100 0 ... 100? Actually, the standard internal ACPI brightness level struct (BRTN in the DSDT) is laid out as: * Full power value (one byte, the value at which the brightness should be set on AC by default) * Economy value (one byte, the value at which the brightness should be set on battery by default) * Actual values (N bytes, up to Max but frequently not) So, no, that value indeed sounds correct. On my laptop the value is: 80 47 0 7 13 20 27 33 40 47 53 60 67 73 80 87 93 100 Thank you for the explanation. FWIW, here is the full output from sysctl hw.acpi.video.lcd0: hw.acpi.video.lcd0.active: 1 hw.acpi.video.lcd0.brightness: 11 hw.acpi.video.lcd0.fullpower: 100 hw.acpi.video.lcd0.economy: 100 hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 hw.acpi.video.crt0.active: 1 hw.acpi.video.ext0.active: 1 hw.acpi.video.ext1.active: 1 hw.acpi.video.ext2.active: 1 hw.acpi.video.ext3.active: 1 hw.acpi.video.ext4.active: 1 hw.acpi.video.ext5.active: 1 What revision of -CURRENT are you running? What is the outcome of trying the patch posted Saturday morning (UTC) from Elizabeth Myers (message ID 54cc5311.9070...@interlinked.me)? I'm currently running r277858 and haven't tried the patch, yet. But will do now and report back. So now I have a new sysctl hw.dri.0.i915_backlight. But no matter what value I pass to it, the brightness doesn't change. It's a Ivy Brigde notebook where drm shows some errors in dmesg. Please see: http://pastie.org/9877917 pgpfUaEvmaEn_.pgp Description: PGP signature
Re: drm2 regression: backlight adjustment on ivybridge no longer works
On 1. Februar 2015 11:20:26 MEZ, Alexandr Krivulya shur...@shurik.kiev.ua wrote: 01.02.2015 12:12, Lars Engels пишет: On Sun, Feb 01, 2015 at 10:41:25AM +0100, Lars Engels wrote: On Sun, Feb 01, 2015 at 03:35:15AM -0600, Andrew Wilcox wrote: Lars Engels sent: 01 February 2015 03:18: With acpi_video I get some interesting sysctl: hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 I guess it should not be 100 100 0 ... 100? Actually, the standard internal ACPI brightness level struct (BRTN in the DSDT) is laid out as: * Full power value (one byte, the value at which the brightness should be set on AC by default) * Economy value (one byte, the value at which the brightness should be set on battery by default) * Actual values (N bytes, up to Max but frequently not) So, no, that value indeed sounds correct. On my laptop the value is: 80 47 0 7 13 20 27 33 40 47 53 60 67 73 80 87 93 100 Thank you for the explanation. FWIW, here is the full output from sysctl hw.acpi.video.lcd0: hw.acpi.video.lcd0.active: 1 hw.acpi.video.lcd0.brightness: 11 hw.acpi.video.lcd0.fullpower: 100 hw.acpi.video.lcd0.economy: 100 hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 hw.acpi.video.crt0.active: 1 hw.acpi.video.ext0.active: 1 hw.acpi.video.ext1.active: 1 hw.acpi.video.ext2.active: 1 hw.acpi.video.ext3.active: 1 hw.acpi.video.ext4.active: 1 hw.acpi.video.ext5.active: 1 What revision of -CURRENT are you running? What is the outcome of trying the patch posted Saturday morning (UTC) from Elizabeth Myers (message ID 54cc5311.9070...@interlinked.me)? I'm currently running r277858 and haven't tried the patch, yet. But will do now and report back. So now I have a new sysctl hw.dri.0.i915_backlight. But no matter what value I pass to it, the brightness doesn't change. It's a Ivy Brigde notebook where drm shows some errors in dmesg. Please see: http://pastie.org/9877917 Please try to boot with acpi_ibm and without acpi_video. This way all works for me. That‘s already what I do. Also tried it with acpi_video without luck. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
On Sat, Jan 31, 2015 at 07:11:23AM +, Miguel Clara wrote: On the laptop I'm running current, if relevant Loading acpi_video doesn't do much I don't see any sysctl related to lcd0 or birightness, this is one of those computers with hybrid graphics, intel + ATI card, so not sure if that's somewhat related. Anyway, I've just tested the drm patches (both) and intelbacklight ( https://github.com/grembo/intel_backlight_fbsd) compiles, installs and runs fine! So thanks both for the patches. On my X230 I also lost the ability to control brightness with Fn+F8 / Fn+F9. With acpi_video I get some interesting sysctl: hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 I guess it should not be 100 100 0 ... 100? Changing brightness with hw.acpi.video.lcd0.brightness doesn't work. pgpimanfO7tsz.pgp Description: PGP signature
Re: drm2 regression: backlight adjustment on ivybridge no longer works
On Sun, Feb 01, 2015 at 03:35:15AM -0600, Andrew Wilcox wrote: Lars Engels sent: 01 February 2015 03:18: With acpi_video I get some interesting sysctl: hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 I guess it should not be 100 100 0 ... 100? Actually, the standard internal ACPI brightness level struct (BRTN in the DSDT) is laid out as: * Full power value (one byte, the value at which the brightness should be set on AC by default) * Economy value (one byte, the value at which the brightness should be set on battery by default) * Actual values (N bytes, up to Max but frequently not) So, no, that value indeed sounds correct. On my laptop the value is: 80 47 0 7 13 20 27 33 40 47 53 60 67 73 80 87 93 100 Thank you for the explanation. FWIW, here is the full output from sysctl hw.acpi.video.lcd0: hw.acpi.video.lcd0.active: 1 hw.acpi.video.lcd0.brightness: 11 hw.acpi.video.lcd0.fullpower: 100 hw.acpi.video.lcd0.economy: 100 hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 hw.acpi.video.crt0.active: 1 hw.acpi.video.ext0.active: 1 hw.acpi.video.ext1.active: 1 hw.acpi.video.ext2.active: 1 hw.acpi.video.ext3.active: 1 hw.acpi.video.ext4.active: 1 hw.acpi.video.ext5.active: 1 What revision of -CURRENT are you running? What is the outcome of trying the patch posted Saturday morning (UTC) from Elizabeth Myers (message ID 54cc5311.9070...@interlinked.me)? I'm currently running r277858 and haven't tried the patch, yet. But will do now and report back. pgpb9IVfFB_TI.pgp Description: PGP signature
RE: drm2 regression: backlight adjustment on ivybridge no longer works
Lars Engels sent: 01 February 2015 03:18: With acpi_video I get some interesting sysctl: hw.acpi.video.lcd0.levels: 100 100 0 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 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 I guess it should not be 100 100 0 ... 100? Actually, the standard internal ACPI brightness level struct (BRTN in the DSDT) is laid out as: * Full power value (one byte, the value at which the brightness should be set on AC by default) * Economy value (one byte, the value at which the brightness should be set on battery by default) * Actual values (N bytes, up to Max but frequently not) So, no, that value indeed sounds correct. On my laptop the value is: 80 47 0 7 13 20 27 33 40 47 53 60 67 73 80 87 93 100 What revision of -CURRENT are you running? What is the outcome of trying the patch posted Saturday morning (UTC) from Elizabeth Myers (message ID 54cc5311.9070...@interlinked.me)? Andrew -- Andrew Wilcox, C/C++/Python developer, kernel hacker Blog: http://blog.foxkit.us/ WWW: http://foxkit.us/ GitHub: https://github.com/awilfox ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
I had to patch manually, for some reason the diff was rejected, but I t compiled and I've instead on 10/stable... problem is this is the laptop where I am typing this :D Its also upgrading some ports, so not a good time to reboot, but I'll post results in the other thread, as for the HP, works perfect (for me at least... far better then what I had :) ) % intel_backlight 50 current backlight value: 100% (4882/4882) set backlight to 50% (2441/4882) % sysctl -a | grep backlig hw.dri.0.i915_backlight: 50 % kldstat |grep acpi 91 0x81fb1000 8708 acpi_video.ko % sudo sysctl hw.dri.0.i915_backlight=100 hw.dri.0.i915_backlight: 50 - 100 % intel_backlight current backlight value: 100% (4882/4882 Melhores Cumprimentos // Best Regards --- *Miguel Clara* *IT - Sys Admin Developer* *E-mail:*miguelmcl...@gmail.com www.linkedin.com/in/miguelmclara/ On Sat, Jan 31, 2015 at 7:40 AM, Elizabeth Myers elizab...@interlinked.me wrote: On 01/31/15 01:28, Miguel Clara wrote: I saw the discussion about how to approach the issue, but totally missed the patch :D Thanks, I'll try that... I'm assuming it will make no difference on 10/stable yet? only current? Excellent question! I have absolutely no idea. I don't even think it applies cleanly. If you can get it to apply and it builds, it just might work. -- Cheers, Elizabeth ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
On 01/31/15 02:00, Miguel Clara wrote: I had to patch manually, for some reason the diff was rejected, but I t compiled and I've instead on 10/stable... problem is this is the laptop where I am typing this :D Its also upgrading some ports, so not a good time to reboot, but I'll post results in the other thread, as for the HP, works perfect (for me at least... far better then what I had :) ) Good to know! If this gets merged (even in modified form), you could ask about an MFC. I'm not the person to ask about this though, I'm just a humble peon. -- Cheers, Elizabeth ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
I saw the discussion about how to approach the issue, but totally missed the patch :D Thanks, I'll try that... I'm assuming it will make no difference on 10/stable yet? only current? On Sat, Jan 31, 2015 at 7:22 AM, Elizabeth Myers elizab...@interlinked.me wrote: On 01/31/15 01:11, Miguel Clara wrote: On the laptop I'm running current, if relevant Loading acpi_video doesn't do much I don't see any sysctl related to lcd0 or birightness, this is one of those computers with hybrid graphics, intel + ATI card, so not sure if that's somewhat related. Anyway, I've just tested the drm patches (both) and intelbacklight ( https://github.com/grembo/intel_backlight_fbsd) compiles, installs and runs fine! So thanks both for the patches. You might be interested in my newest patch to add backlight control to sysctl. It's probably better for everyone (imo) that backlight control is integrated into the driver and accessible via sysctl, so it works the same way for everyone. Look for the subject Re: Questions on adding backlight support for the i915 driver from me and you'll find the patch. I am curious if it works for you. -- Cheers, Elizabeth ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
On the laptop I'm running current, if relevant Loading acpi_video doesn't do much I don't see any sysctl related to lcd0 or birightness, this is one of those computers with hybrid graphics, intel + ATI card, so not sure if that's somewhat related. Anyway, I've just tested the drm patches (both) and intelbacklight ( https://github.com/grembo/intel_backlight_fbsd) compiles, installs and runs fine! So thanks both for the patches. ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
On 01/31/15 01:11, Miguel Clara wrote: On the laptop I'm running current, if relevant Loading acpi_video doesn't do much I don't see any sysctl related to lcd0 or birightness, this is one of those computers with hybrid graphics, intel + ATI card, so not sure if that's somewhat related. Anyway, I've just tested the drm patches (both) and intelbacklight ( https://github.com/grembo/intel_backlight_fbsd) compiles, installs and runs fine! So thanks both for the patches. You might be interested in my newest patch to add backlight control to sysctl. It's probably better for everyone (imo) that backlight control is integrated into the driver and accessible via sysctl, so it works the same way for everyone. Look for the subject Re: Questions on adding backlight support for the i915 driver from me and you'll find the patch. I am curious if it works for you. -- Cheers, Elizabeth ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
On 01/31/15 01:28, Miguel Clara wrote: I saw the discussion about how to approach the issue, but totally missed the patch :D Thanks, I'll try that... I'm assuming it will make no difference on 10/stable yet? only current? Excellent question! I have absolutely no idea. I don't even think it applies cleanly. If you can get it to apply and it builds, it just might work. -- Cheers, Elizabeth ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
2015-01-28 10:11 GMT+01:00 Adrian Chadd adr...@freebsd.org: ugh, so because i can't sleep, I read the Linux i915 history for backlight changes. Turns out if I disable bit 30 in ivb_pch_pwm_override() to register BLC_PWM_PCH_CTL1, the acpi video setting works. The keyboard settings still don't, but I'll worry about that later. Bit 30 in the linux drm code is: #define BLM_PCH_OVERRIDE_ENABLE (1 30) The linux dri code doesn't set that for ivybridge - only for broadwell. (also, the linux driver has POSTING_READ() which I'm guessing is to ensure things made it out to the device and have been flushed before doing another IO operation.) I have my (via sysctl) panel backlight control working again. -adrian Thank you guys, the backlight control works again on my laptop with this patch, suggested by Adrian : --- sys/dev/drm2/i915/intel_display.c.orig (revision 277487) +++ sys/dev/drm2/i915/intel_display.c(working copy) @@ -6995,7 +6995,7 @@ */ I915_WRITE(BLC_PWM_CPU_CTL2, PWM_ENABLE); I915_WRITE(BLC_PWM_CPU_CTL, 0); -I915_WRITE(BLC_PWM_PCH_CTL1, PWM_ENABLE | (130)); +I915_WRITE(BLC_PWM_PCH_CTL1, PWM_ENABLE); } void intel_modeset_init_hw(struct drm_device *dev) but the patch by Andrew does not hurts: http://foxkit.us/FreeBSD/i915-uninitialised-var-fix.diff I am running r277871 with both patches applied. Regards, Maurizio ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
29.01.2015 12:46, Ranjan1018 . пишет: 2015-01-28 10:11 GMT+01:00 Adrian Chadd adr...@freebsd.org: ugh, so because i can't sleep, I read the Linux i915 history for backlight changes. Turns out if I disable bit 30 in ivb_pch_pwm_override() to register BLC_PWM_PCH_CTL1, the acpi video setting works. The keyboard settings still don't, but I'll worry about that later. Bit 30 in the linux drm code is: #define BLM_PCH_OVERRIDE_ENABLE (1 30) The linux dri code doesn't set that for ivybridge - only for broadwell. (also, the linux driver has POSTING_READ() which I'm guessing is to ensure things made it out to the device and have been flushed before doing another IO operation.) I have my (via sysctl) panel backlight control working again. -adrian Thank you guys, the backlight control works again on my laptop with this patch, suggested by Adrian : --- sys/dev/drm2/i915/intel_display.c.orig (revision 277487) +++ sys/dev/drm2/i915/intel_display.c(working copy) @@ -6995,7 +6995,7 @@ */ I915_WRITE(BLC_PWM_CPU_CTL2, PWM_ENABLE); I915_WRITE(BLC_PWM_CPU_CTL, 0); -I915_WRITE(BLC_PWM_PCH_CTL1, PWM_ENABLE | (130)); +I915_WRITE(BLC_PWM_PCH_CTL1, PWM_ENABLE); } void intel_modeset_init_hw(struct drm_device *dev) This patch fixes problem for me too, but only when acpi_video module is not loaded. With acpi_video loaded I still see this errors (but it's probably another topic): Jan 29 19:02:27 thinkpad kernel: ACPI Exception: AE_AML_PACKAGE_LIMIT, Index (0x000FF) is beyond end of object (length 0x16) (20140926/exoparg2-452) Jan 29 19:02:27 thinkpad kernel: ACPI Error: Method parse/execution failed [\_SB_.PCI0.LPCB.EC0_.BRNS] (Node 0xf80007541b00), AE_AML_PACKAGE_LIMIT (20140926/psparse-552) Jan 29 19:02:27 thinkpad kernel: ACPI Error: Method parse/execution failed [\_SB_.PCI0.VID_.LCD0._BCM] (Node 0xf80007550280), AE_AML_PACKAGE_LIMIT (20140926/psparse-552) Jan 29 19:02:27 thinkpad kernel: can't evaluate \_SB_.PCI0.VID_.LCD0._BCM - AE_AML_PACKAGE_LIMIT ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
ugh, so because i can't sleep, I read the Linux i915 history for backlight changes. Turns out if I disable bit 30 in ivb_pch_pwm_override() to register BLC_PWM_PCH_CTL1, the acpi video setting works. The keyboard settings still don't, but I'll worry about that later. Bit 30 in the linux drm code is: #define BLM_PCH_OVERRIDE_ENABLE (1 30) The linux dri code doesn't set that for ivybridge - only for broadwell. (also, the linux driver has POSTING_READ() which I'm guessing is to ensure things made it out to the device and have been flushed before doing another IO operation.) I have my (via sysctl) panel backlight control working again. -adrian ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
Hi! Good catch! But nope, didn't fix it. -adrian On 27 January 2015 at 23:23, Andrew Wilcox awil...@wilcox-tech.com wrote: Ranjan1018 . sent: 27 January 2015 23:42: The regression was introduced in r277487. Looking closer at this revision, I believe I may see the issue. Most of the new tunables introduced in dev/drm2/i915/i915_drv.c are uninitialised and thus probably contain junk. It likely works for me because the firmware of my laptop 0s memory on EFI boot. I have uploaded a diff here that I believe may fix this. Please test and let me know: http://foxkit.us/FreeBSD/i915-uninitialised-var-fix.diff Best, Andrew -- Andrew Wilcox, C/C++/Python developer, kernel hacker Blog: http://blog.foxkit.us/ WWW: http://foxkit.us/ GitHub: https://github.com/awilfox ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
and the commit from linux: commit 7aa1e7f06d6ea1bce3b27630d50769d13da28b1a Author: Daniel Vetter daniel.vet...@ffwll.ch Date: Wed Jun 27 16:43:36 2012 +0200 Revert drm/i915: allow PCH PWM override on IVB This reverts commit f82cfb6bcda164ef3a66b8c3fc549b1f9bdd09ad. This breaks the backlight controls on my IVB asus zenbook with an eDP panel. I guess the right fix would be to read this bit and use either the pch or the cpu register to frob the backlight values. But that is stuff for -next. Cc: Jesse Barnes jbar...@virtuousgeek.org Signed-Off-by: Daniel Vetter daniel.vet...@ffwll.ch On 28 January 2015 at 01:11, Adrian Chadd adr...@freebsd.org wrote: ugh, so because i can't sleep, I read the Linux i915 history for backlight changes. Turns out if I disable bit 30 in ivb_pch_pwm_override() to register BLC_PWM_PCH_CTL1, the acpi video setting works. The keyboard settings still don't, but I'll worry about that later. Bit 30 in the linux drm code is: #define BLM_PCH_OVERRIDE_ENABLE (1 30) The linux dri code doesn't set that for ivybridge - only for broadwell. (also, the linux driver has POSTING_READ() which I'm guessing is to ensure things made it out to the device and have been flushed before doing another IO operation.) I have my (via sysctl) panel backlight control working again. -adrian ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
RE: drm2 regression: backlight adjustment on ivybridge no longer works
Ranjan1018 . sent: 27 January 2015 23:42: The regression was introduced in r277487. Looking closer at this revision, I believe I may see the issue. Most of the new tunables introduced in dev/drm2/i915/i915_drv.c are uninitialised and thus probably contain junk. It likely works for me because the firmware of my laptop 0s memory on EFI boot. I have uploaded a diff here that I believe may fix this. Please test and let me know: http://foxkit.us/FreeBSD/i915-uninitialised-var-fix.diff Best, Andrew -- Andrew Wilcox, C/C++/Python developer, kernel hacker Blog: http://blog.foxkit.us/ WWW: http://foxkit.us/ GitHub: https://github.com/awilfox ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
RE: drm2 regression: backlight adjustment on ivybridge no longer works
Ranjan1018 . sent: 27 January 2015 23:42: my Samsung laptop has an Intel IvyBridge: [snip] The regression was introduced in r277487. The backlight adjustment works in FreeBSD 11.0-CURRENT r277395, r277486 but not in r277487, r277534 and r277639. Hrm. That is interesting. A few questions I have then. - What happens if you set drm.i915.invert_brightness to -1 in kenv(1) or /boot/loader.conf? - What happens if you set drm.i915.invert_brightness to 1 in kenv(1) or /boot/loader.conf? - What version of graphics/libdrm do you have installed? Did you rebuild it after installing the new kernel? (This shouldn't be necessary, but I am trying to gather all details.) Let's start there and see if we can pin down a cause. Best, Andrew -- Andrew Wilcox, C/C++/Python developer, kernel hacker Blog: http://blog.foxkit.us/ WWW: http://foxkit.us/ GitHub: https://github.com/awilfox ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
2015-01-27 7:38 GMT+01:00 Andrew Wilcox awil...@wilcox-tech.com: Ranjan1018 . sent: 26 January 2015 06:19: 2015-01-24 23:25 GMT+01:00 Adrian Chadd adr...@freebsd.org: The backlight adjustment doesn't work on my ivybridge mobile laptop (Lenovo X230) after the dri update. I have the same issue on my Samsung Ativ 2 laptop. I have a Sandy Bridge laptop (Apple MacBook Pro 8,2) - HD 3000: vgapci0@pci0:0:2:0: class=0x03 card=0x00db106b chip=0x01268086 rev=0x09 hdr=0x00 vendor = 'Intel Corporation' device = '2nd Generation Core Processor Family Integrated Graphics Controller' class = display subclass = VGA It is running: FreeBSD pwyll.foxkit.us 11.0-CURRENT #1 r277781M: Mon Jan 26 18:41:15 CST 2015 r...@pwyll.foxkit.us:/usr/obj/usr/src/sys/GENERIC amd64 I have no issues using acpi_video's sysctls (hw.acpi.video.lcd0.brightness) to adjust backlight, though it does not have good granularity. The stepping is about 7, so it goes as 13%..20%..27%..35%..etc. The intel_backlight program from intel-gpu-tools also no longer changes the backlight value. This program works fine for me on both an older kernel (r277523) and this kernel (r277781), after applying some patches to allow the library to build on FreeBSD. It also has better granularity (the stepping is 2-3). If there is anything I can do/run to aide in debugging why it works for me but not others, let me know. Best, Andrew -- Andrew Wilcox, C/C++/Python developer, kernel hacker Blog: http://blog.foxkit.us/ WWW: http://foxkit.us/ GitHub: https://github.com/awilfox Hi, my Samsung laptop has an Intel IvyBridge: vgapci0@pci0:0:2:0: class=0x03 card=0xc708144d chip=0x01668086 rev=0x09 hdr=0x00 vendor = 'Intel Corporation' device = '3rd Gen Core processor Graphics Controller' class = display subclass = VGA The regression was introduced in r277487. The backlight adjustment works in FreeBSD 11.0-CURRENT r277395, r277486 but not in r277487, r277534 and r277639. Regards, Maurizio ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
Re: drm2 regression: backlight adjustment on ivybridge no longer works
2015-01-24 23:25 GMT+01:00 Adrian Chadd adr...@freebsd.org: Hi, I've just found a regression. The backlight adjustment doesn't work on my ivybridge mobile laptop (Lenovo X230) after the dri update. I've added debugging. It's making it all the way to the pch backlight panel update routine in intel_panel.c, but it's not changing the backlight appearance itself. The intel_backlight program from intel-gpu-tools also no longer changes the backlight value. I'm going to finish rebuilding -HEAD on the sandy bridge laptop here and try it out. Thanks, -adrian Hi, I have the same issue on my Samsung Ativ 2 laptop. Regards, Maurizio ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
RE: drm2 regression: backlight adjustment on ivybridge no longer works
Ranjan1018 . sent: 26 January 2015 06:19: 2015-01-24 23:25 GMT+01:00 Adrian Chadd adr...@freebsd.org: The backlight adjustment doesn't work on my ivybridge mobile laptop (Lenovo X230) after the dri update. I have the same issue on my Samsung Ativ 2 laptop. I have a Sandy Bridge laptop (Apple MacBook Pro 8,2) - HD 3000: vgapci0@pci0:0:2:0: class=0x03 card=0x00db106b chip=0x01268086 rev=0x09 hdr=0x00 vendor = 'Intel Corporation' device = '2nd Generation Core Processor Family Integrated Graphics Controller' class = display subclass = VGA It is running: FreeBSD pwyll.foxkit.us 11.0-CURRENT #1 r277781M: Mon Jan 26 18:41:15 CST 2015 r...@pwyll.foxkit.us:/usr/obj/usr/src/sys/GENERIC amd64 I have no issues using acpi_video's sysctls (hw.acpi.video.lcd0.brightness) to adjust backlight, though it does not have good granularity. The stepping is about 7, so it goes as 13%..20%..27%..35%..etc. The intel_backlight program from intel-gpu-tools also no longer changes the backlight value. This program works fine for me on both an older kernel (r277523) and this kernel (r277781), after applying some patches to allow the library to build on FreeBSD. It also has better granularity (the stepping is 2-3). If there is anything I can do/run to aide in debugging why it works for me but not others, let me know. Best, Andrew -- Andrew Wilcox, C/C++/Python developer, kernel hacker Blog: http://blog.foxkit.us/ WWW: http://foxkit.us/ GitHub: https://github.com/awilfox ___ freebsd-current@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org