#asahi · 2022-10-26 13:57
boriskmarcan: FYI, I've tested similar (but no identical) C++ compilation setups on Mac OS vs Asahi Linux and under Mac OS the build is about 15% slower. About the setups: both using Clang with libc++ but different versions of Clang (14 vs ~13ish) and different targets/libc (Linux vs Darwin). #asahi · 2022-10-25 15:41
boriskOverall, M1 Max has way, way better and faster memroy/cache subsystem and it's made on a better node, so I sort of expected it will at least match i9. Also, this is subjective (can't measure power draw), but i9 doesn't seem to run especially hot: the fan is not audible and the exhaust air is barely warm, about on par with Mac Studio. #asahi · 2022-10-25 15:36
boriskj`ey: both have 64GB. Specifically, DDR5-4400 with ECC in the i9 case and LPDDR5-6400 in the M1 Max case.#asahi · 2022-10-25 15:34
borisk_jannau_: this is native compilation on both. Yes, I realize I am not comparing the same workloads, especially if optimization is involved (since it's optimizing for different architectures). Though I see the same difference with debug builds.#asahi · 2022-10-25 15:32
boriskmarcan: ok, thanks for confirming this is not unexpected#asahi · 2022-10-25 14:43
boriskmarcan, jannau: thanks for the feedback. I should have mentioned M1 Max is in Mac Studio, not a laptop. The numbers I get is that M1 Max is about ~30% slower, depending which compiler I use (GCC or Clang). Might be a good idea to try Apple Clang under Mac OS to make sure there are no drastic discrepancies.#asahi · 2022-10-25 14:00
boriskmarcan: Would there be any reason to think that M1 Max under Asahi is not running at full throttle? I am running the 6.0rc6 kernel and the C++ compilation performance is surprisingly underwhelming compared to i9-12900K, especially considering the differences in the memory bandwidth (which C++ compilation is quite sensitive to, IMO).#asahi · 2022-10-15 15:17
boriskBTW, there appears to be a speed difference between back and front ports (perhaps unsurprisingly). If I plug a USBC-NVME drive in front (USB-C), I get "SuperSpeed Plus Gen 2x1" and if at the back (USB-A) I get just "SuperSpeed".#asahi · 2022-10-15 15:05
boriskthanks a lot for you help, much appreciated!#asahi · 2022-10-15 15:02
boriskanything else you want me to try?#asahi · 2022-10-15 15:01
boriskalso I tried warm reboot, and that works as well, at least from successful previous boot#asahi · 2022-10-15 14:58
boriskyes, seems to work, devices plugged into both front and back ports are detected. Also the same Serial/Manufacturer as with the manual test.#asahi · 2022-10-15 14:55
boriskbooting...#asahi · 2022-10-15 14:51
boriskthe DT changes compiled ok with GPIO_PULL_UP#asahi · 2022-10-15 14:41
boriskok#asahi · 2022-10-15 14:33
boriskok, will try now#asahi · 2022-10-15 14:28
boriskcan retest to confirm if this is important#asahi · 2022-10-15 14:28
boriskyes, you could be right, but I believe the error was different#asahi · 2022-10-15 14:26
boriskok, if you want me to test updated DT to confirm, I can easily do that#asahi · 2022-10-15 14:25
boriskAlso, before I forgot: I had to use DTs bundled with u-boot from asahi-6.0-rc6-1 kernel. With DTs from 5.19 the asmedia firmware would not load with 6.0 kernel.#asahi · 2022-10-15 14:23
boriskok, makes sense. Do you want me to file an issue with m1n1 to capture this?#asahi · 2022-10-15 14:20
boriskYes, before 1850->in when I had the hub but no alive ports, there was no serial and the manufacturer was GenesisLogic. Now I have serial and the manufacturer changes to Apple. Even the number of ports changes. Bizarre...#asahi · 2022-10-15 14:17
borisk"SerialNumber: 7423J07"#asahi · 2022-10-15 14:14
boriskdo you want me to try with 1849?#asahi · 2022-10-15 14:13
borisk1851->0, 1850->in, 1851->1 -- both front and back ports are a live now#asahi · 2022-10-15 14:09
borisksorry, what does "1850 to input" mean? Just read it?#asahi · 2022-10-15 14:06
boriskwrite 1 to both see what happens?#asahi · 2022-10-15 14:04
borisk(I've written 1 back to 1851 and got the hub back, BTW)#asahi · 2022-10-15 14:04
borisk1849: out/0#asahi · 2022-10-15 14:03
borisk1850: out/0#asahi · 2022-10-15 14:02
boriskyes, writing 0 to 1851 I get "USB disconnected"#asahi · 2022-10-15 14:00
boriskmaybe set 1851 back to 0?#asahi · 2022-10-15 13:58
boriskdirection in, value 1#asahi · 2022-10-15 13:54
boriskyes, I think the "no power" theory could be correct: when I plug into these ports, the led on the device does not come on (I've tried two different)#asahi · 2022-10-15 13:52
boriskmarcan: but the ports are still dead, both front and back#asahi · 2022-10-15 13:49
boriskyes, doing that I get a bunch of kernel messages about new USB3.1 Hub (Manufacturer: GenesysLogic)#asahi · 2022-10-15 13:45
boriskvalue: 0#asahi · 2022-10-15 13:45
boriskdirection: out#asahi · 2022-10-15 13:43
boriskyes, cold boot helps, trying gpio stuff now#asahi · 2022-10-15 13:40
borisk(that's before messing with gpio)#asahi · 2022-10-15 13:40
boriskuh, strange, now I am getting "Firmware version is too old after upload"#asahi · 2022-10-15 13:35
boriskjannau: the lsusb output is the same with and without asmedia firmware loaded#asahi · 2022-10-15 13:34
boriskok, sec, let me re-enable the firmware loading#asahi · 2022-10-15 13:32
borisk1620#asahi · 2022-10-15 13:32
boriskyes, I have this symlink#asahi · 2022-10-15 13:30
boriskmarcan: yes, I do#asahi · 2022-10-15 13:28
borisk(I have that part nailed so it might be quicker than setting up m1n1 proxy)#asahi · 2022-10-15 13:27
boriskmarcan: got it, if you can propose a code fragment to add to m1n1 (presumably stage 2), I can give it a try#asahi · 2022-10-15 13:24
boriskmarcan: I am way out of my depth here, but shoudn't this GPIO be poked after loading the firmware (which happens in the kernel)?#asahi · 2022-10-15 13:22
boriskmarcan: not at the moment but I can probably set it up