ChanServ changed the topic of #linux-msm to:
marvin24 has joined #linux-msm
marvin24_ has quit [Ping timeout: 480 seconds]
<sboyd> DavidHeidelberg[m]: I think sic means secondary interrupt controller
<sboyd> there was certainly a time when they jammed various "apps" bits into one block, i.e. the CPU clock control and how it pinged other processors in the SoC were all combined into one hardware block because it was sorta related
cxl000 has quit [Remote host closed the connection]
cxl000 has joined #linux-msm
cxl000 has quit [Remote host closed the connection]
cxl000 has joined #linux-msm
<DavidHeidelberg[m]> sboyd: Hey, do you still recall clk framework bit? We use APQ8064 (at least for Nexus 7, for others it seems to work without it) https://github.com/apq8064-mainline/linux/commit/8d364f90d16aa0b0c967d97015b8b65fa6583362 . I added some debug printf to your patch https://paste.sr.ht/blob/660f9380ed0f66b704737a9647389e56d1374aa3 and on the similar apq8064 devices output https://www.toptal.com/developers/hastebin/raw/uboyimeteh any
<DavidHeidelberg[m]> ideas, if it could be solvable differently?
Rayyan is now known as Rayyan|alt
Rayyan has joined #linux-msm
lumag_ has quit [Remote host closed the connection]
lumag_ has joined #linux-msm
<DavidHeidelberg[m]> bamse: "Can't we just pick one of them?" you mean ordering (ref, sleep or sleep, ref) ? In case we would pick one of them, we could pick two sleep or two ref values
<DavidHeidelberg[m]> one thing I looking for is to find a way, to define -> these values, but I don't care about the order
<bamse> DavidHeidelberg[m]: i was thinking just to say ref, sleep...
<DavidHeidelberg[m]> bamse: and change it in DTS files?
<bamse> right
<DavidHeidelberg[m]> yeah, sounds good for me
<bamse> but if you don't want to care about the order...can't you use allOf?
<bamse> i mean, i do think you should just pick ref,sleep and fix the dts...but i'm curious
<DavidHeidelberg[m]> yeah, I think I don't have nice answer for this, since `allOf: enum [ref, sleep]` is just an one value
<DavidHeidelberg[m]> I have some weird feeling I have seen it solved somewhere, but I don't think any possible combination didn't worked for me
<DavidHeidelberg[m]> ok, I was just too lazy.
<Marijn[m]> David Heidelberg: Wasn't it:... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/jSqlTOUZJXzipOJLLTpJVbfr)
<DavidHeidelberg[m]> the order is the issue
<DavidHeidelberg[m]> but... maxItems: 2; contains: items ....
<DavidHeidelberg[m]> Marijn: beware, this is primary IRC room, they'll scream about multilines
<Marijn[m]> I've sent multiline text here before... Does it spit it out as multiple messages, even the `` ``` `` markers?
<DavidHeidelberg[m]> `maxItems:2\n contains:\n items:\n - const: a\n - const: b`
<DavidHeidelberg[m]> that's my current solution which works
<minecrell> 19:40 <Marijn[m]> David Heidelberg: Wasn't it:... (full message at https://matrix.org/_matrix/media/r0/download/matrix.org/jSqlTOUZJXzipOJLLTpJVbfr)
<DavidHeidelberg[m]> validates number of entries, and check each in the list is there
<DavidHeidelberg[m]> minecrell: I hope in faster Matrix adoption :D
<Marijn[m]> Gah, /me goes back to 1988 :P
<DavidHeidelberg[m]> linux-msm was one of channels, who for long time had ##linux-msm which was kinda impossible enter with many IRC clients, so maybe get back pre 1988 :D
<Marijn[m]> David Heidelberg: could you substitute the `items: ...` with `enum: [a, b]` too?
<Marijn[m]> As long as it resides in a `contains:`, that is
<DavidHeidelberg[m]> I'll test, as long it removes const: and simplifies it, then it's good
<Marijn[m]> Afaik `enum:` is supposed to be synonymous to something with `oneOf` and a list of `const`
<Marijn[m]> Does `items:\n - const: a\n - const: b` work? Seems like that also requires the ordering to be fixed, or is that solved through `contains:`?
<Marijn[m]> `example-schema.yaml` has exactly one use of `contains:` and doesn't seem to explain what it does :(
<Marijn[m]> `arm/cpus.yaml` has an example of that `contains: enum: - a - b` structure
<DavidHeidelberg[m]> I don't like using existing yamls as docs.. but I do that.. 😆
<DavidHeidelberg[m]> yeah, but that does validate only one entry
<DavidHeidelberg[m]> anyway, in this case, it seems to iterate over both, since: a) it can have duplicates (so no = "sleep", "sleep") b) maxItems: 2
<DavidHeidelberg[m]> *can't
<DavidHeidelberg[m]> btw. are u in #devicetree channel? We are kinda unrelated to MSM now :D
<Marijn[m]> It's sad that there's no proper, easy-to-use representation for the prevalent `clocks = <&p>, <&handle>` and their `clock-names = "ref", "sleep"` pattern
<Marijn[m]> I can be :)
gpiccoli_ has joined #linux-msm
gpiccoli has quit [Remote host closed the connection]