dshankar has quit [Quit: dshankar]
dshankar has joined #imx6-dev
dshankar has quit [Ping timeout: 260 seconds]
dshankar has joined #imx6-dev
dshankar has quit [Quit: dshankar]
cnxsoft has joined #imx6-dev
dshankar has joined #imx6-dev
dshankar has quit [Quit: dshankar]
eebrah has joined #imx6-dev
eebrah is now known as Guest48450
Guest48450 is now known as eebrah_
eebrah_ has quit [Read error: Connection reset by peer]
eebrah_ has joined #imx6-dev
eebrah_ has quit [Ping timeout: 240 seconds]
cnxsoft has quit [Ping timeout: 245 seconds]
cnxsoft has joined #imx6-dev
diego_r has joined #imx6-dev
dshankar has joined #imx6-dev
abrasive has quit [Read error: Operation timed out]
abrasive has joined #imx6-dev
eebrah_ has joined #imx6-dev
cnxsoft has quit [Ping timeout: 245 seconds]
cnxsoft has joined #imx6-dev
fossxplorer has joined #imx6-dev
JCxMLnblFl has joined #imx6-dev
JCxMLnblFl has left #imx6-dev [#imx6-dev]
<dshankar> looks while messing with aptitude i broke everything :-) it's now just a console boot.
rz2k has joined #imx6-dev
rz2k has quit []
eebrah_ has quit [Ping timeout: 248 seconds]
dshankar has quit [Quit: dshankar]
rz2k has joined #imx6-dev
abrasive has quit [Remote host closed the connection]
abrasive has joined #imx6-dev
eebrah_ has joined #imx6-dev
cnxsoft has quit [Quit: cnxsoft]
<JBD1986> all, is there a list of required defconfig/.config items for the vivante drivers to work properly?
jas-hacks has joined #imx6-dev
<jas-hacks> JBD1986: I tested on the debian rootfs again, I don't see any issues while dragging a windows around. The only issue I'm seeing is when a window is resized there a delay repainted and the CPU jumps to 60 %.
diego_r has quit [Ping timeout: 272 seconds]
<JBD1986> jas-hacks, when you very rapidly move a window up and down on top of the desktop icons, do the desktop icons redraw instantly, or is there a delay?
<jas-hacks> JBD1986: icon or any windows behind is drawn instantly
<JBD1986> that sounds like my performance with compositing enabled
<JBD1986> but with compositing disabled, redraw is much faster, and moving is slow, and redraw of icons is delayed
<JBD1986> sorry resize is much faster
<JBD1986> not redraw
<JBD1986> compositing= fast window move, fast icon redraw, SUPER slow resizing
<JBD1986> no compositing= laggy window movement, slow icon redraw. Window resize is reasonable, but still not great
<jas-hacks> yes that matches
<JBD1986> wait, so if you turn off compositing, your window movement is laggy?
<JBD1986> and icon redraw is laggy?
<jas-hacks> yes it is not as smooth with it off, which makes sense compositing should improve how the window stack is handled
<JBD1986> jas-hacks, I thought you had compositing disabled and did not experience window moving lag?
<JBD1986> high cpu utilization of xorg
<jas-hacks> with or without it compositing I have issues with window moving lag
<jas-hacks> /s/issues/no issues
<JBD1986> jas-hacks, so with compositing disabled, you can wildly drag the window around the screen, and icons refresh properly, and the window is not lagging behind your mouse?
<jas-hacks> the only symptom is slow icon redraw
<WarheadsSE> hmm
<jas-hacks> JBD1986: looks like it is because no h/w acceleration is used for icon/windows in the background because they aren't in off screen memory
<JBD1986> jas-hacks, but there is 0 lag when dragging windows with compositing off, while moving a large window rapidly?
<JBD1986> in terms of the window following your mouse
<JBD1986> ?
<jas-hacks> I don't see that
<JBD1986> jas-hacks, and cpu usage stays normal for xorg?
<JBD1986> i.e. less than 50% of a single core?
<jas-hacks> it only goes up if I wildly drag the window but thats to be expected
<newell> Anyone in here ever use the global timer for the i.MX6?
<jas-hacks> newell: I haven't
<JBD1986> jas-hacks, does the issue get noticeably worse if the window takes up a large portion of the screen?
<JBD1986> jas-hacks, simply moving a large window from the bottom right corner to the top left corner of my screen has a noticeable lag
<newell> jas-hacks, have you used any of the timers on the i.MX6? We are trying to use one of the timers for a 125 microsecond interrupt to measure the interrupt latency of the system.
<newell> I am getting hung up on how exactly to go about doing that
<newell> We want to isolate one of the cores as well...uncertain on how I would do that as well.
<jas-hacks> newell: nothing in the SDK?
<newell> Those are baremetal programs...and we would like to use this within Linux
<jas-hacks> JBD1986: I don't see much difference
<newell> So I am thinking I will need to write a Linux kernel driver to do this just uncertain of how to go about it.
eebrah_ has quit [Ping timeout: 260 seconds]
<rz2k> other ones are either used internally by other imx/mxc drivers or not available, but you can try to use them since you have all the definitions
<WarheadsSE> jas-hacks: we tried going to a full 6q & 1G ram, no help.
<WarheadsSE> I actually have one of the more-production models here, that is seeing a vertical tearing @ window drag
<rz2k> linux kernel has a framework for defining timers and various thing around them.
<JBD1986> jas-hacks, hste, WarheadsSE is refering to my issue
<JBD1986> CPU utilization of xorg shoots to 24%, which is about all of a single core
<JBD1986> jas-hacks, you are testing in 1080p correct?
<newell> rz2k, so is your suggestion to look at that link and then base my driver off of that?
<rz2k> newell: my suggestion is to read https://www.kernel.org/doc/Documentation/timers/ :p
<rz2k> and define everything you need yourself
<newell> rz2k, thanks :)
<WarheadsSE> jas-hacks, all: [drm] Initialized vivante 1.0.0 20120216 on minor 0 = correct version?
<jas-hacks> WarheadsSE: yes correct version
<JBD1986> jas-hacks, how does this line read in your xorg log?
<JBD1986> II) VIVANTE(0): hardware: DISP3 BG (video memory: 8100kB)
<jas-hacks> I'm running at 1280x1024 so video memory: 5120kB
<jas-hacks> JBD1986: what your seeing is CPU rendering
<JBD1986> jas-hacks, when you tested the window drag, we're you at 1080p or 1280x1024?
<JBD1986> were*
<jas-hacks> 1280x1024
<jas-hacks> Given the x viv extensions only support:
<jas-hacks> [ 19.855] (II) Solid
<jas-hacks> [ 19.855] (II) Copy
<jas-hacks> [ 19.855] (II) Composite (RENDER acceleration)
<jas-hacks> [ 19.855] (II) UploadToScreen
<jas-hacks> not sure what your expecting
<JBD1986> jas-hacks, was looking for matchin performance with you
<JBD1986> jas-hacks, it still lags at 1280x1024 as well
<JBD1986> just tested
eebrah_ has joined #imx6-dev
<WarheadsSE> jas-hacks: what Hz?
<jas-hacks> 60Hz, that should make no difference
<WarheadsSE> K
<WarheadsSE> we've tried 1080p@60, 1280x1024@60, all see the same problems
<jas-hacks> WarheadsSE: which rootfs?
<WarheadsSE> ours and yours
<jas-hacks> how bad is the tearing?
<JBD1986> jas-hacks, when moving windows in a diagonal, it is a staircase effect
<WarheadsSE> the size of the steps width is expounded by speed
<jas-hacks> you see it on the scrollbar?
<WarheadsSE> the whole frame of the window on the verticl edges
<jas-hacks> ok I see it with XTerm
<JBD1986> jas-hacks, have you tested the 2D tests that get installed with the 2D library?
<WarheadsSE> jas-hacks: yeah, we're getting it with xterm, glxgears (and the es gears), the xfce file manager, etc.
<JBD1986> jas-hacks, I tried all of your terminal programs as well
<JBD1986> jas-hacks, not just xterm
<jas-hacks> not sure if it is tearing or just a slow repaint
<JBD1986> jas-hacks, have you tested the 2D tests that get installed with the 2D library?
<JBD1986> /opt/fsl-samples/
staylor has joined #imx6-dev
<jas-hacks> JBD1986: i did run the 2d tests
<JBD1986> jas-hacks, do you see anything when you run /opt/fsl-samples/g2d/g2d_test?
<JBD1986> i get console output, but I don't see anything
<jas-hacks> they are unit tests
<jas-hacks> with no visible screen output
<JBD1986> I do get an overlay from the overlay test
<JBD1986> visibly
<JBD1986> but nothing for the g2d_test
<JBD1986> jas-hacks, do you know if there is any documentation for those unit tests?
<jas-hacks> its not a problem with 2d libraries, its more do with the x extension support from vivante
<jas-hacks> if you compare the xorg.log from your PC with one from your board you will notice the support is minimal
jas-hacks has left #imx6-dev [#imx6-dev]
eebrah_ has quit [Ping timeout: 272 seconds]
eebrah_ has joined #imx6-dev
dshankar has joined #imx6-dev
eebrah_ has quit [Ping timeout: 248 seconds]
rz2k has quit []
bfederau has quit [Read error: Connection reset by peer]
bfederau has joined #imx6-dev
projectgus has quit [Ping timeout: 252 seconds]
<newell> anyone here know how to get freescale's 3.10.17 kernel for yocto (I am using Freescale's github code).
<newell> ?
<newell> Looks like the newer kernel uses a device tree while 3.0.35 doesn't and I would like to upgrade to this kernel version
<newell> otavio, ping
<otavio> newell: pong
<otavio> newell: using master-next and PREFERRED_VERSION
<newell> thx otavio :)