#
✕ nick:rqou_
#asahi-re · 2022-04-07 18:17
rqou_the typedef in the KDK for "dart_blk_sid_config_v8_t" has a bunch of disable bits, e.g. "disable_ttbr_invalid_err" "disable_ste_invalid_error" etc.
#asahi-re · 2022-04-07 18:15
rqou_also, from the symbols, t6000 doesn't have the error disable bits that t8020 has in the TCR (none of which are in m1n1)
#asahi-re · 2022-04-07 18:13
rqou_i'm driving them with m1n1 python scripts and polling for the done bit
#asahi-re · 2022-04-07 18:13
rqou_i do know of such devices (jpeg, prores), but i don't know how to get any useful timings
#asahi-re · 2022-04-07 18:11
rqou_t8020 also appears to have wrprot at bit7 and rdprot at bit8 according to the symbols
#asahi-re · 2022-04-07 18:10
rqou_i'm not sure how i would go about testing whether device accesses are bypassing cache though
#asahi-re · 2022-04-07 18:10
rqou_which for t6000 and t8110 indeed look the same, with bits 0-3 being valid, uncacheable, wrprot, rdprot
#asahi-re · 2022-04-07 18:09
rqou_sven: re DART "uncacheable": i'm going off of the DWARF symbols in the KDK KernelSupport object files which contain a DW_TAG_typedef for "tXXXXdart_l3tte_t"
#asahi-re · 2022-04-07 06:06
rqou_is this related to AGX compression or is this a different way to make data be tiled?
#asahi-re · 2022-04-07 06:06
rqou_does anybody know what IOSurface::getTileHeaderRegionOffsetOfPlane does?
#asahi-re · 2022-04-07 03:44
rqou_i asuume _that_ might be related to some of the magic flags that are also not understood
#asahi-re · 2022-04-07 03:42
rqou_in addition to the magic quality parameters, i do still have a concern of "how the heck does the hardware know what size of descriptor it should be reading?"
#asahi-re · 2022-04-07 03:26
rqou_in other news, i have an experiment that's able to drive the prores encoder usefully
#asahi-re · 2022-04-07 03:26
rqou_anybody have any guesses what these prores parameters might control? https://github.com/rqou/m1n1/blob/prores/proxyclient/experiments/prores.py#L621
#asahi-re · 2022-04-06 11:08
rqou_works well enough to encode a single frame with hardcoded settings
#asahi-re · 2022-04-06 11:08
rqou_wip reverse engineering of the prores encoder https://github.com/rqou/m1n1/tree/prores
#asahi-re · 2022-04-03 11:05
rqou_not sure i even own any thunderbolt devices
#asahi-re · 2022-04-03 11:04
rqou_ok, i'll look at this some other time
#asahi-re · 2022-04-03 11:04
rqou_F
#asahi-re · 2022-04-03 11:03
rqou_even if i just want to read register 0?
#asahi-re · 2022-04-03 10:33
rqou_(i'm just trying to check the parameter registers to see if they're very different or not)
#asahi-re · 2022-04-03 10:32
rqou_hmm, then i get 0xffffffff trying to read any of the DART registers
#asahi-re · 2022-04-03 10:22
rqou_0xffffffffffffffff
#asahi-re · 2022-04-03 10:22
rqou_TTY> pmgr: Error getting /arm-io/dart-apciec0 clock-gates.
#asahi-re · 2022-04-03 10:22
rqou_>>> p.pmgr_adt_clocks_enable(f'/arm-io/dart-apciec0')
#asahi-re · 2022-04-03 10:22
rqou_btw, why doesn't attempting to turn on dart-apcieN work?
#asahi-re · 2022-04-03 07:50
rqou_idk if it bypasses all caches or just the iommu TLB
#asahi-re · 2022-04-03 07:49
rqou_this only affects iommu pages, but yes
#asahi-re · 2022-04-03 05:58
rqou_unrelatedly, does anybody know how to fix ghidra's varargs ABI?
#asahi-re · 2022-04-03 05:57
rqou_at some point should investigate whether or not that bit actually makes the page not be cached (presumably in the DART TLB?)
#asahi-re · 2022-04-03 05:57
rqou_idk, i'm not really a good kernel dev
#asahi-re · 2022-04-03 05:40
rqou_idk if the bit actually works
#asahi-re · 2022-04-03 05:40
rqou_Someone(TM) should maybe investigate/fix that?
#asahi-re · 2022-04-03 05:40
rqou_ah
#asahi-re · 2022-04-03 05:37
rqou_uhh where? can we make it not do that?
#asahi-re · 2022-04-03 05:13
rqou_according to the symbols, it flags the page as uncacheable
#asahi-re · 2022-04-03 05:13
rqou_uhh the linux DART driver doesn't try to set PTE bit 1 (SP_PROT_DIS) does it? because on t6000 that bit does something completely different
#asahi-re · 2022-04-03 00:41
rqou_possibly even requiring a PPL bypass
#asahi-re · 2022-04-03 00:41
rqou_unless someone wants to hand me a research jailbreak on an a15
#asahi-re · 2022-04-03 00:41
rqou_oh, the code is available, just the hardware behavior wouldn't be
#asahi-re · 2022-04-03 00:40
rqou_mix of both
#asahi-re · 2022-04-03 00:38
rqou_note that this is hardware revision 1.0, and the actual M2 may have yet another new version of this DART interface
#asahi-re · 2022-04-03 00:37
rqou_i think i have the core of the t8110 dart figured out https://github.com/rqou/m1n1/tree/dart-t8110
#asahi-re · 2022-04-02 21:49
rqou_there should be DART performance counters that can be queried if we want to investigate this in depth
#asahi-re · 2022-04-02 21:46
rqou_based on debug strings in the driver that mention "client partitions"
#asahi-re · 2022-04-02 21:45
rqou_i suspect it might affect the way the TLB is used/divided up
#asahi-re · 2022-04-02 21:45
rqou_hmm i had a sudden inspiration/guess as to what DART tunables might affect, but i'm not sure how we could test that
#asahi-re · 2022-04-01 21:11
rqou_does anybody have any idea what a "CTC" might be in the context of the DART?
#asahi-re · 2022-03-28 10:19
rqou_from back when they pwned the broadcom wifi
#asahi-re · 2022-03-28 10:19
rqou_of the DART

plus ›