#
✕ nick:kettenis
#asahi · 2021-06-03 16:20
kettenisright
#asahi · 2021-06-03 16:19
kettenisthe assiocated DART has more features though it seems
#asahi · 2021-06-03 16:18
kettenisonly has a sinle root port
#asahi · 2021-06-03 16:18
kettenisjudging from the corellium code it is largely the same
#asahi · 2021-06-03 16:14
kettenisand everthing is built-in, so no slots where folks can plug in PCIe bridges
#asahi · 2021-06-03 16:14
kettenisin the end for the current hardware it doesn't matter all that much since we have separate DARTs for each root port and only one device on each root port
#asahi · 2021-06-03 16:12
kettenisyeah
#asahi · 2021-06-03 16:12
kettenisI mapped everything to SID 1 since SID 0 seems to be a catch-all for all RIDs that aren't mapped explicitly in the PCIe host bridge
#asahi · 2021-06-03 16:11
kettenisdo you have any thoughts on the PCI RID to DART SID mapping we should use?
#asahi · 2021-06-03 16:10
kettenismaking the clocks optional in u-boot fixed that issue
#asahi · 2021-06-03 16:10
kettenismy problem was related to the commented out clocks, which as a cascading failure resulted in the pci bus numbering being wrong which in turn screwed up the iommu mapping
#asahi · 2021-06-03 16:09
kettenissven: I've added the DART interrupts to u-boot
#asahi · 2021-06-03 16:06
kettenisthe one-clock-per-node model still has my preference
#asahi · 2021-06-03 16:03
ketteniswe deliberately don't want to have to encode the relationships in the driver
#asahi · 2021-06-03 16:02
kettenissven: yeah, no clock peoplehanging out here it seems
#asahi · 2021-06-03 15:58
ketteniscould be a simple array of clock numbers
#asahi · 2021-06-03 15:57
kettenisfor example "clock-parents" or maybe "apple,clock-parents"
#asahi · 2021-06-03 15:57
kettenisI wonder if you should just use a different name for that clocks property
#asahi · 2021-06-03 15:49
kettenisnot much automatic clock processing u-boot
#asahi · 2021-06-03 15:49
kettenisbut I think it should be fine
#asahi · 2021-06-03 15:48
ketteniswill have to check u-boot
#asahi · 2021-06-03 15:42
kettenisthe self-referencing clock property doesn't trigger any magic within the linux clock framework?
#asahi · 2021-06-03 15:41
kettenisyeah, that looks somewhat reasonable
#asahi · 2021-06-03 09:56
ketteniscorrect
#asahi · 2021-06-03 09:54
kettenisthat is my understanding
#asahi · 2021-06-03 09:37
kettenisor we could add invent properties that have different names but work in a similar way
#asahi · 2021-06-03 09:34
kettenisbut if the people that matter think it is fine, then I won't object
#asahi · 2021-06-03 09:33
kettenissven: I believe assigned-clocks isn't the appropriate mechanism
#asahi · 2021-06-02 21:31
kettenisyeah, so I wonder what is going wrong there...
#asahi · 2021-06-02 21:30
kettenisI'll see if I can deal with that tomorrow
#asahi · 2021-06-02 21:28
kettenisheh, well, it causes dart interrupts to fire on the openbsd side ;)
#asahi · 2021-06-02 21:23
kettenisI can probably pick up thhose minimal changes to the u-boot dt
#asahi · 2021-06-02 21:20
kettenisbut let's see what the reaction is on your dart patches before moving stuff around
#asahi · 2021-06-02 21:19
kettenisavoiding the top as well because that is where the framebuffer is
#asahi · 2021-06-02 21:18
kettenisI picked something a bit higher up since low memory is used by the firmware and not available
#asahi · 2021-06-02 21:17
kettenisonly 512MB but that can be easily changed
#asahi · 2021-06-02 21:17
kettenisI think I'm currently mapping 0x960000000-0x980000000
#asahi · 2021-06-02 21:14
kettenisbut we'll need to agree on the exact range that we map
#asahi · 2021-06-02 21:14
kettenisthe dma-ranges hack might actually help me with u-boot as well
#asahi · 2021-06-02 21:12
kettenisah, yes, of course
#asahi · 2021-06-02 21:09
kettenisu-boot needs the clocks though
#asahi · 2021-06-02 21:07
kettenisonce we've sorted the clock and dart bindings the u-boot device tree should just work
#asahi · 2021-06-02 21:05
kettenisso linux completely ignores the dma-ranges proprty if there is an enabled iommu?
#asahi · 2021-06-02 21:03
kettenisright, that's what I do for u-boot
#asahi · 2021-06-02 21:02
kettenisbut bypass doesn't work for the pcie darts is it?
#asahi · 2021-06-02 21:00
kettenislast version removes the "phy" set of registers, so that will require an adjustment in the driver code
#asahi · 2021-06-02 21:00
kettenisit seems robh is mostly happy with my pcie dt binding
#asahi · 2021-06-02 20:58
kettenisAFAIK linux network drivers poll
#asahi · 2021-06-02 20:56
kettenisah, right
#asahi · 2021-06-02 20:53
kettenisso using stream ID #1 works fine with your code?

plus ›