1 NOTES about msm drm/kms driver:
3 In the current snapdragon SoC's, we have (at least) 3 different
4 display controller blocks at play:
5 + MDP3 - ?? seems to be what is on geeksphone peak device
6 + MDP4 - S3 (APQ8060, touchpad), S4-pro (APQ8064, nexus4 & ifc6410)
7 + MDP5 - snapdragon 800
9 (I don't have a completely clear picture on which display controller
12 Plus a handful of blocks around them for HDMI/DSI/etc output.
14 And on gpu side of things:
15 + zero, one, or two 2d cores (z180)
16 + and either a2xx or a3xx 3d core.
18 But, HDMI/DSI/etc blocks seem like they can be shared across multiple
19 display controller blocks. And I for sure don't want to have to deal
20 with N different kms devices from xf86-video-freedreno. Plus, it
21 seems like we can do some clever tricks like use GPU to trigger
22 pageflip after rendering completes (ie. have the kms/crtc code build
23 up gpu cmdstream to update scanout and write FLUSH register after).
25 So, the approach is one drm driver, with some modularity. Different
26 'struct msm_kms' implementations, depending on display controller.
27 And one or more 'struct msm_gpu' for the various different gpu sub-
30 (Second part is not implemented yet. So far this is just basic KMS
31 driver, and not exposing any custom ioctls to userspace for now.)
33 The kms module provides the plane, crtc, and encoder objects, and
34 loads whatever connectors are appropriate.
36 For MDP4, the mapping is:
38 plane -> PIPE{RGBn,VGn} \
39 crtc -> OVLP{n} + DMA{P,S,E} (??) |-> MDP "device"
40 encoder -> DTV/LCDC/DSI (within MDP4) /
41 connector -> HDMI/DSI/etc --> other device(s)
43 Since the irq's that drm core mostly cares about are vblank/framedone,
44 we'll let msm_mdp4_kms provide the irq install/uninstall/etc functions
45 and treat the MDP4 block's irq as "the" irq. Even though the connectors
46 may have their own irqs which they install themselves. For this reason
47 the display controller is the "master" device.
49 For MDP5, the mapping is:
51 plane -> PIPE{RGBn,VIGn} \
52 crtc -> LM (layer mixer) |-> MDP "device"
54 connector -> HDMI/DSI/eDP/etc --> other device(s)
56 Unlike MDP4, it appears we can get by with a single encoder, rather
57 than needing a different implementation for DTV, DSI, etc. (Ie. the
58 register interface is same, just different bases.)
60 Also unlike MDP4, with MDP5 all the IRQs for other blocks (HDMI, DSI,
61 etc) are routed through MDP.
63 And finally, MDP5 has this "Shared Memory Pool" (called "SMP"), from
64 which blocks need to be allocated to the active pipes based on fetch
67 Each connector probably ends up being a separate device, just for the
68 logistics of finding/mapping io region, irq, etc. Idealy we would
69 have a better way than just stashing the platform device in a global
70 (ie. like DT super-node.. but I don't have any snapdragon hw yet that
73 Note that so far I've not been able to get any docs on the hw, and it
74 seems that access to such docs would prevent me from working on the
75 freedreno gallium driver. So there may be some mistakes in register
76 names (I had to invent a few, since no sufficient hint was given in
77 the downstream android fbdev driver), bitfield sizes, etc. My current
78 state of understanding the registers is given in the envytools rnndb
81 https://github.com/freedreno/envytools/tree/master/rnndb
82 (the mdp4/hdmi/dsi directories)
84 These files are used both for a parser tool (in the same tree) to
85 parse logged register reads/writes (both from downstream android fbdev
86 driver, and this driver with register logging enabled), as well as to
87 generate the register level headers.