Mirrors/m1n1
2
0
Fork 0
mirror of https://github.com/AsahiLinux/m1n1 synced 2024-12-25 04:43:04 +00:00
Code Issues Projects Releases Packages Wiki Activity
2150 commits 13 branches 49 tags 9 MiB
Commit graph

1349 commits

Author SHA1 Message Date
Hector Martin
9b7ae1b443 m1n1.fw.mtp: Fix parsing of init messages 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-16 21:23:46 +09:00
Hector Martin
36bcc36173 m1n1.fw.asc.oslog: Implement properly 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-16 21:23:46 +09:00
Hector Martin
d5267123c1 experiments/mtp.py: Clean up 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-16 21:23:46 +09:00
Asahi Lina
19e00f1f23 m1n1.hv: Remove __OS_LOG segments from coprocessors 
These are supposed to be after the kernel, so we'd have to move them
for the guest to be able to access them. It seems things work if we just
delete those segments.

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-04-16 21:23:46 +09:00
Janne Grunau
22890f3ba1 experiments: Optimize bad apple video playback 
Let ffmpeg rotate, align to stride and convert to RGBA.
Do not allocate a framebuffer for every frame.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-04-16 17:11:33 +09:00
Sasha Finkelstein
5681036b2e Add touchbar screen experiments 
Signed-off-by: Sasha Finkelstein <fnkl.kernel@gmail.com>
 2023-04-15 17:23:36 +09:00
Hector Martin
ad53766d22 m1n1.fw.asc.{crash, ioreporting}: Do not align prealloc buffer sizes 
On SMC on t602x, this is a small SRAM and we can run off the edge.

Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-12 00:40:06 +09:00
Hector Martin
aaeb3000e6 m1n1.hv: t6020 support 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-12 00:40:06 +09:00
Hector Martin
3bc3b0131f tools/pmgr_adt2dt.py: Add multidie support 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-09 23:50:42 +09:00
Hector Martin
06884b5613 experiments/cpu_pstate_latencies.py: Fix SoC configs 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-09 21:04:40 +09:00
Hector Martin
2389fa9d3d m1n1.adt: Fix parsing of template ADTs 
Adding the speaker calibration stuff broke it because we try to parse
template values as real values. Don't do that.

Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-09 19:46:13 +09:00
Hector Martin
fe104d3848 m1n1.hv: Also hook ATC_AON device 
It seems 13.2 is messing with this now.

Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-09 19:46:13 +09:00
Hector Martin
795211c534 m1n1.fw.dcp.ipc: More fields 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-04-09 19:46:13 +09:00
Mario Hros
693ebbae2b m1n1.hv: Support T6021 cpustart offset 
Signed-off-by: Mario Hros <git@reversity.org>
 2023-04-09 19:46:13 +09:00
Mario Hros
69012c0702 experiments/cpu_pstate_latencies.py: Extend for T6021 
Signed-off-by: Mario Hros <git@reversity.org>
 2023-04-09 19:46:12 +09:00
Asahi Lina
8bf08b0ff1 linux.py: Implement TSO 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-28 19:06:20 +09:00
Asahi Lina
0a05a0171e proxy: Add smp_is_alive() thunk 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-28 19:06:20 +09:00
Eileen Yoon
6ce14d8735 m1n1/ane: Initial commit 
Signed-off-by: Eileen Yoon <eyn@gmx.com>
 2023-03-28 17:10:15 +09:00
Hector Martin
8be0596ae8 experiments/mtp.py: Fix DART stuff 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-03-15 13:54:35 +09:00
Hector Martin
43a737b608 m1n1.adt: Add speaker calibration parsing 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-03-11 20:09:57 +09:00
Asahi Lina
7ad099a5b6 m1n1.trace.agx: Add exploit mitigation 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
6d3e594903 m1n1.hv: Fix read-only or write-only hooks 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
5cacbdc501 m1n1.fw.agx.initdata: Define buffer manager ctl fields better 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
90f4957725 experiments/agx_xtest.py: Here be exploits! 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
3548d4d210 m1n1.fw.ags.microsequence: Add more ops 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
e8377889d2 m1n1.proxyutils: Add 13.0 beta5/6 versions 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
c35168bb88 m1n1.hw.uat: Gracefully handle inaccessible page tables 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
b063fde19a m1n1.hw.uat: Support block maps 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
bfaf358f82 m1n1.agx.render: Add microsequence hooks 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
e49c90a314 m1n1.agx.context: Add GPUMicroSequence.cur_addr() 
This requires that we pre-allocate the buffer

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
586be52ec5 m1n1.trace.agx: Log structure UAT permissions as meta 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
b4c687ce13 m1n1.hw.uat: Add ioperm() 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
cc1284f264 hv/trace_agx_defer.py: Exclude context ID 1 (compositor) 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
9a8da80e8d m1n1.trace.agx: Add exclude_context_id option 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
9e93bd70e4 m1n1.fw.agx: More compute and blit stuff 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
074892377c m1n1.fw.agx: Fix some more sequence number-ish field names 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
b7fb8d98d8 m1n1.agx: Increase available VM size and reduce usage 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
25813d91b4 m1n1.trace.agx: Disable RegionC tracing 
The atomic stuff breaks with Linux right now...

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
3e572aba4a m1n1.fw.dcp.ipc: Add compression info struct 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
813c545509 m1n1.fw.agx: Compute struct fixes for 13.2 
Assuming these apply at 13.0b4 too...

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
f01a62b7a1 m1n1.fw.agx.cmdqueue: Misc fixes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
1400b50157 m1n1.fw.agx: Initial 13.2 structure changes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
e22d2fe049 m1n1.hw.agx: Fix fault status register fields 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
e7dee84c3a m1n1.trace.agx: Misc TA/3D dump changes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
7e49d104ea m1n1.trace.agx: Add compute tracing 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
8fba6118ce m1n1.fw.agx: Add missing compute stuff 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
89e3fcb2ab m1n1.trace.agx: Fix delayed ring buffer fetches 
This should solve the issue where we see commands that the GPU has
already processed.

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
37f9c5149c m1n1.proxyutils/constructutils: Add missing versions 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
3117cdef99 m1n1.fw.agx: Fixes for 13.0b4 support 
Found while adding 13.2, but almost certainly mistakes from back when we
did 13.0b4.

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Asahi Lina
91e0809063 m1n1.fw.agx: Misc initdata/etc fixes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-03-06 15:15:17 +09:00
Janne Grunau
5f3bd50b4a experiments/dart_dump.py: Fix register dump after 183991ca19 
Fixes: 183991ca19 ("m1n1.hw.dart: Hide all DART variants behind common interface")
Signed-off-by: Janne Grunau <j@jannau.net>
 2023-02-02 09:50:56 +09:00
Janne Grunau
2b4ed4abbf hv/trace_cd3217: Add tracing for TI/Apple cd3217 USB type C controller 
Interface (as used by macOS) is simple enough to skip register
definitions.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:49:04 +09:00
Janne Grunau
6f4c53279e hv/trace_dcp.py: Add support for for M2 
Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:48:46 +09:00
Janne Grunau
aaa04cfaba m1n1.trace.asc: Support tracing of a DART SID != 0 
Required for DCP on M2.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:48:46 +09:00
Janne Grunau
f8964ed0f9 m1n1.fw.asc: Increase DVA size to 36-bits 
A single TTBR covers 36-bit address space with 16k pages. Necessary for
DCP on M2.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:48:46 +09:00
Janne Grunau
39f933ec1f m1n1.fw.dcp.dcpep: DCPEp_SetShmem allows to query Shmem 
When used with FLAG = 0, IOMFB replies with the current Shmem IOVA. It
is NULL on the first call and keeps the last set value. Updating it
seems to work without issues after shutdown.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:48:46 +09:00
Janne Grunau
5631aaa74a experiments/scaler.py: Use unified DART implementation 
Untested if that makes this work on M1 systems which use dart8020 for
the scaler.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:48:46 +09:00
Janne Grunau
68bba35263 experiments/aes.py: Simplify by using unified DART 
Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:48:46 +09:00
Janne Grunau
183991ca19 m1n1.hw.dart: Hide all DART variants behind common interface 
Allows m1n1 experiments and tracers not to care about the DART variant.
Required to trace and experiment with DCP sanely on M1 and M2.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:48:46 +09:00
Janne Grunau
1fc26dbb1a hv/trace_dcp: Disable Disp0/DCP iboot EP tracing 
Its usage in m1n1 it is incompatible with the generic EPICEp tracing.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:48:46 +09:00
Janne Grunau
bf73c1416a m1n1.fw.asc.ioreporting: Support IOP provided buffers 
SMC provides buffer from its MMIO space. Since the macOS 13 SMC firmware
the firmware crashes when ACK-ing a GetBuf which provided a buffer.

Signed-off-by: Janne Grunau <j@jannau.net>
 2023-01-28 10:46:45 +09:00
Hector Martin
46f2811351 m1n1.fw.dcp.iboot: Identify colorspace values 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2023-01-22 23:57:41 +09:00
Asahi Lina
3d28ac4d6f tools/chainload.py: Properly sleep DCP when chainloading macOS 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-01-20 18:18:13 +09:00
Asahi Lina
7a92257d30 m1n1.agx.render: Change heap size to 0x200 
This seems to be the AGX value

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-01-20 18:18:13 +09:00
Asahi Lina
e774b5c521 m1n1.trace.agx: Log hook writes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-01-20 18:18:13 +09:00
Asahi Lina
8e77f8f49f m1n1.fw.agx.channels: Do not require contiguous pages for channel state 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2023-01-20 18:18:13 +09:00
Martin Povišer
2cb6cb46ca tools/run_guest.py: Add -v option for attaching 9P virtio exports 
Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Martin Povišer
93c20e9395 m1n1.hv: Allocate virtio resources if unspecified 
Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Martin Povišer
020d86a23f m1n1.utils: Fix len() overflow in RangeMap 
This is to fix

  OverflowError: Python int too large to convert to C ssize_t

in case the upper bound of a range is 1<<64.

Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Martin Povišer
c760e02388 hv: Insert attached virtios into ADT 
Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Martin Povišer
d94f918433 m1n1.adt: Add to_bus_addr() method 
So that one knows what to put in reg properties.

Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Martin Povišer
6f460b1f6a m1n1.adt: Add create_node() method 
Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Martin Povišer
2b33d7f801 m1n1.adt: Implement 'in' keyword on ADT nodes 
So that one can check for children, like:

  >>> "/arm-io/aic" in u.adt
  True

Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Martin Povišer
0249d04a4b m1n1.adt: Add missing 'sys' import 
Used for sys.stderr later on.

Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Martin Povišer
6661ab14d0 hv: Add 9P virtio peripheral 
Add some minimal implementation of virtio peripherals. At the level
of on-target hypervisor code we implement the MMIO layout and
maintain virtqueues. Once a buffer is available, we break into the
host proxyclient to deal with it.

Specific device-classes of the virtio spec ought to be implemented in
the proxyclient. Here the one device implemented is 9P transport,
exporting the m1n1 source directory.

Signed-off-by: Martin Povišer <povik@cutebit.org>
 2023-01-13 17:25:25 +09:00
Hector Martin
586157b8bf experiments/smc_watcher.py: new experiment 
Signed-off-by: Hector Martin <marcan@marcan.st>
 2022-12-26 15:07:52 +09:00
Sasha Finkelstein
8c2ad9ada6 Add z2 touch panel tracing 
Signed-off-by: Sasha Finkelstein <fnkl.kernel@gmail.com>
 2022-12-24 20:45:19 +09:00
Asahi Lina
87fa247d9b m1n1.fw.agx.microsequence: Identify visibility_result_buffer field 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
9ebc38ca73 m1n1.trace.agx: Fix UAT TTBR PT size 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
1b9cb35104 experiments/agx_*.py: Autodetect FW version and GPU rev 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
2b0e022b78 m1n1.agx: Port rendering to G14 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
696cdee114 m1n1.agx.initdata: Port to G14G 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
80266ff93e m1n1.constructutils: Unify version conditional syntax with Rust 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
82a9bf3c26 hv/trace_agx_pwr.py: Cleanup 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
9360a92cdc m1n1/trace/agx.py: Add SGXTracer 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
0a57f45805 m1n1.hw.agx: Add chip info regs 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
e1752306c5 m1n1.fw.agx.initdata: Fix a bunch of fields 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
2dab8e352a m1n1.agx.render: Misc tiling related fixes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
9e65bd527b experiments/agx_{parallel,renderframe}.py: Misc fixes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
25e1644858 m1n1.adt: Add mtr-polynom-fuse-agx parsing 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
fbc04995ab tools/chainload.py: Fix chainloading of xnu 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
6e99fde8f2 hv/trace_all_more.py: Trace more stuff 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
10d772f743 hv/trace_agx_pwr.py: Messing with pstates 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
34b883318c m1n1.fw.agx.initdata: Make somethings Dec 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
e4810a452d m1n1.utils: Add align_pot() 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
7b63613945 m1n1.agx.render: Always enable Z compression 
This seems to be required on t600x...

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
05ece42eba m1n1.agx.render: Better t600x support 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
bcd3d0274d m1n1.agx.initdata: Work out all the ADT relationships 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
d265abc90c m1n1.adt: Fixes for adding properties 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
808baf22df m1n1.agx: Initial T600x port 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:10 +09:00
Asahi Lina
5a73bd6476 hv/trace_agx_pwr.py: New HV experiment 
This is intended to reverse-engineer the macOS driver's power scale
calculations.

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:17:07 +09:00
Asahi Lina
02d9fdc6a0 m1n1.trace.agx: Add after_init_hook 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
e50be4bf0d experiments/agx_tlb.py: Misc tracing changes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
29ee0d5d32 m1n1.hv: Don't catch SystemExit from tracers/shells 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
b4d5ff7520 m1n1.agx: Add printing IRQs 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
2fe37d6dfa m1n1.fw.agx: Misc field renames & fixes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
b4a19572f2 m1n1.fw.agx.handoff: Add CUR_CTX field 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
3aa2be6489 m1n1.hw.uat: Flushing cleanup 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
f92be5a076 experiments/agx_parallel.py: Uncomment some power-related tracers 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
f6dad68156 m1n1.fw.agx.initdata: Disable KTrace by default 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
20fce92280 m1n1.trace.agx: Disable some hex dumps 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
b8bf3f485e m1n1.trace.agx: Various UAT related improvements 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
a097e5b744 m1n1.trace.agx: Instantiate SGXRegs 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
621e7b535a m1n1.fw.agx: Misc structure tweaks 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
2978f246f2 m1n1.agx.render: PowerVR style tile param calculation 
Also document some other registers

Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
c72b6ce98c m1n1.hv: Translate DC CIVAC argument if trapped 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
633e551c2d m1n1.agx.object: Support reading back directly from physmem 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
bc99d3411a m1n1.agx.context: Fix blocklist indexing 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
5e39db820b m1n1.agx: Only try to decode fault if fault happened 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
1f28498b0a m1n1.hw.agx: Export R_FAULT_INFO 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
b8f9008953 m1n1.hw.uat: Show context ID in exceptions 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
b04cc0c7c8 m1n1.hw.uat: Document attribute indexes 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Asahi Lina
6a0008884d m1n1.agx: Add an initdata hook 
Signed-off-by: Asahi Lina <lina@asahilina.net>
 2022-12-04 13:15:55 +09:00
Hector Martin
9e2eeeb911 hv/trace_pmgr.py: Get rid of pwrgate stuff 
It's broken

Signed-off-by: Hector Martin <marcan@marcan.st>
 2022-12-04 13:10:59 +09:00
Janne Grunau
71baff4adc experiments/dcp.py: Set the display with surface swap 
Needs testing on DP/HDMI outputs.

Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Janne Grunau
7d9608adc4 m1n1.fw.dcp.manager: Add backlight callbacks 
DCP reports the display brightness with
PropRelay::pr_publish(prop_id=15, value=...). The value divided by
"Brightness_scale" appears to be the display brightness in nits.
Add other callbacks encountered on j314c.

Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Janne Grunau
6aa1da3c31 m1n1.fw.dcp.ipc: Annotate backlight fields in swap_rec 
Allows brightness control of integrated displays.

Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Janne Grunau
98e8e48171 experiments/dcp.py: Add set_digital_out_mode call 
Initializes the display if not already done by iboot/m1n1. Not expected
to change anything for disp0 but might be helful in bringing DP alt
mode for dispexp[0-7].

Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Janne Grunau
053eb7bfda experiments/dcp.py: Handle external displays correctly 
Pass disp0's comaptible string to DCPManager.
Fixes Mac Studio and M1 Macbook handling.

Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Janne Grunau
452a3243c0 m1n1.fw.dcp.manager: Add SetDigitalOutMode convenience function 
This prints the selected modes. In addition omits printing the whole
DCPAVProps.

Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Janne Grunau
dc3397888d m1n1.fw.dcp.manager: Add update_backlight_factor_prop call back 
Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Janne Grunau
aa7088f3fb m1n1.fw.dcp.manager: Support 't600x' in rt_bandwidth_setup_ap 
Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Janne Grunau
e91cc63708 m1n1.fw.dcp.manager: Fix swap_complete_intent_gated after fb49d88d90 
Signed-off-by: Janne Grunau <j@jannau.net>
 2022-11-29 01:36:14 +09:00
Martin Povišer
de0c0323a5 experiments/mmio_sweep.py: Patch in AOP power-domains on t6000 
The ADT lacks some of the info on AOP power domains on t6000, so fill
that in manually.

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
fc66046d76 experiments/aop.py: Extend AOP experiment, add tracer 
The AOP uses an 'EPIC' protocol similar to the one other coprocessor
firmware is using but not in the exact same version. Add code for
tracing the AOP calls and extend the aop.py experiment with the client
side of it. Include description of audio calls and some other calls
related to sensor discovery.

Furthermore, in experiments/aop.py, do some AOP audio setup. Once that
is done we can start streaming samples from the internal microphones by
making what AOP considers power state adjustment calls. That is, we
adjust the power state of a 'hpai' device, first to a 'pw1 ' stage,
then to 'pwrd' stage.

So, to see microphone samples, enter the AOP experiment shell first:

  $ M1N1DEVICE=/dev/ttyACM0 experiments/aop.py

Within the shell, adjust the power state of 'hpai':

  >>> aop_set_audio_pstate('hpai', 'pw1 ')

At that point /arm-io/admac-aop-audio powers up. In parallel to the AOP
shell, we can start tools/admac_stream.py on the just powered-up ADMAC
instance:

  $ M1N1HEAP=0x10010000000 M1N1DEVICE=/dev/ttyACM1 tools/admac_stream.py \
        --node admac-aop-audio --channel 1 -v | xxd -g 4 -c 12 -e

Returning back to the AOP shell, we can then set 'hpai' to 'pwrd' state
to kick off the streaming:

  >>> aop_set_audio_pstate('hpai', 'pwrd')

By that point, we should see samples coming out on the ADMAC end. The
samples are 32-bit floats packed in groups of three in a frame, e.g.

00000000: ba7ac6a7 ba32d3c3 baa17ae2  ..z...2..z..
0000000c: 38ccea5f b99c1a37 ba0c4bb1  _..87....K..
00000018: 39d2354f 3964b5ff 39b209fb  O5.9..d9...9
00000024: b96a1d1f 39c8503f 3958fc4f  ..j.?P.9O.X9
00000030: b6b1f5ff 39c72b8f 39bbe017  .....+.9...9
0000003c: 3a912de5 36dd4f7f 37f1147f  .-.:.O.6...7

This has been tested and will to some degree be specific to 2021 Macbook
Pro (t6000). Differences on other models TBD (at the very least the
number of microphones can be presumed different).

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
a8ef753520 m1n1.fw.aop: In AOPBase only refer to ProxyUtils attribute 
Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
abd0c2ac3c m1n1.fw.aop: Tweak bootargs diff format 
Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
2bacc044e8 m1n1.fw.aop: Move bootargs helpers to new class AOPBase 
The new class can possibly be shared between the m1n1 AOP tracer and
AOP client.

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
6ad59beb0c experiments/aop.py: Allocate oslog buffer on request 
We need to allocate a buffer for the AOP on the OSLog endpoint for it to
fully boot. Copy in a modified version of the general OSLog endpoint
driver to do that.

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
f87d65a971 experiments/aop.py: Make _bootargs_span a property 
Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
1d2a157fe8 m1n1.fw.aop: Split off ASCArgumentSection from experiments/aop.py 
Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
79890a15d6 experiments/aop.py: Work around ADT inconsistencies 
In t6000 ADT, the AOP SRAM base is specified *including* the bus offset
where the bus offset isn't expected, so our decoding of it includes the
bus offset twice. Patch it.

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
b6ae985377 m1n1.fw.afk: Accept AOP buffer magic 
Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
8347301eff m1n1.hw.adt: Add prop parsing rules related to AOP audio 
We need to adapt the prop-parsing logic to give precedence to path
matching over name matching, since otherwise the audio-* rule (which is
already in the list) catches (...)/aop-audio/audio-pdm2. We want that
to fall under *aop-audio* rule instead.

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
fcf6170a9f m1n1.proxyutils: Accept M1N1HEAP environment variable 
Let user override the proxyutils heap base for concurrent running of two
proxyclient sessions.

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
d735d6ad49 m1n1.hw.dart: Pass in kwargs arguments of from_adt() 
Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
9a93d5c307 m1n1.trace: Drop stale super().__init__() call 
The Reloadable parent doesn't have an __init__ method, so calling the
super may call at unexpected places in richer inheritance hierarchies.

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
c5e80f0d5a m1n1.fw.afk: Fix TX/RX_BUFSIZE confusion 
Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
ce517b10a3 tools/admac_stream.py: Add tool for engaging ADMAC 
Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00
Martin Povišer
8ff430aefa m1n1.hw.admac: Add CHAN_SRAM_CARVEOUT registers 
The new registers contain SRAM limits allocated to each channel.

Signed-off-by: Martin Povišer <povik@protonmail.com>
 2022-11-04 12:36:49 +09:00