#asahi-dev on 2025-09-04 — irc logs at oftc.catirclogs.org

00:04 hdbngr has quit [Ping timeout: 480 seconds]

00:06 zumi_ has joined #asahi-dev

00:08 OriginalEnthusiast has quit [Remote host closed the connection]

00:10 zumi has quit [Ping timeout: 480 seconds]

00:11 zumi_ is now known as zumi

00:11 hdbngr has joined #asahi-dev

00:12 psykose_ has quit [Remote host closed the connection]

00:15 psykose has joined #asahi-dev

00:28 hdbngr has quit [Ping timeout: 480 seconds]

00:36 hdbngr has joined #asahi-dev

00:38 nst has quit [Ping timeout: 480 seconds]

00:47 nst has joined #asahi-dev

01:17 Halian has quit [Quit: I'll be back.]

01:38 hdbngr has quit [Ping timeout: 480 seconds]

01:45 hdbngr has joined #asahi-dev

01:54 hdbngr has quit [Ping timeout: 480 seconds]

01:56 hdbngr has joined #asahi-dev

02:05 hdbngr has quit [Ping timeout: 480 seconds]

02:12 hdbngr has joined #asahi-dev

02:14 hexdump02 has quit [Ping timeout: 480 seconds]

02:21 hdbngr has quit [Ping timeout: 480 seconds]

02:28 hdbngr has joined #asahi-dev

02:37 hdbngr has quit [Ping timeout: 480 seconds]

02:37 tobhe_ has joined #asahi-dev

02:44 tobhe has quit [Ping timeout: 480 seconds]

02:47 hdbngr has joined #asahi-dev

02:56 hdbngr has quit [Ping timeout: 480 seconds]

03:03 skipwich_ has quit [Remote host closed the connection]

03:03 skipwich has joined #asahi-dev

03:04 hdbngr has joined #asahi-dev

03:07 creddy has joined #asahi-dev

03:13 hdbngr has quit [Ping timeout: 480 seconds]

03:21 hdbngr has joined #asahi-dev

03:30 hdbngr has quit [Ping timeout: 480 seconds]

03:30 creddy has quit [Remote host closed the connection]

03:38 hdbngr has joined #asahi-dev

03:47 hdbngr has quit [Ping timeout: 480 seconds]

03:54 hdbngr has joined #asahi-dev

03:58 nora has joined #asahi-dev

04:00 nora_ has quit [Ping timeout: 480 seconds]

04:03 hdbngr has quit [Ping timeout: 480 seconds]

04:10 kevans91_ has quit [Read error: Connection reset by peer]

04:10 kevans91 has joined #asahi-dev

04:11 hdbngr has joined #asahi-dev

04:17 john-cabaj has quit [Ping timeout: 480 seconds]

04:19 hdbngr has quit [Ping timeout: 480 seconds]

04:28 hdbngr has joined #asahi-dev

04:29 eluks has quit [Remote host closed the connection]

04:30 eluks has joined #asahi-dev

04:30 cylm has joined #asahi-dev

04:43 hdbngr has quit [Ping timeout: 480 seconds]

04:50 hdbngr has joined #asahi-dev

04:54 kevans91 has quit [Remote host closed the connection]

04:59 hdbngr has quit [Ping timeout: 480 seconds]

05:00 hdbngr has joined #asahi-dev

05:03 kevans91 has joined #asahi-dev

05:13 hdbngr has quit [Ping timeout: 480 seconds]

05:15 kaazoo has quit [Quit: Leaving.]

05:21 hdbngr has joined #asahi-dev

05:21 kaazoo has joined #asahi-dev

05:33 hdbngr has quit [Ping timeout: 480 seconds]

05:40 hdbngr has joined #asahi-dev

05:49 hdbngr has quit [Ping timeout: 480 seconds]

05:51 kaazoo has quit [Quit: Leaving.]

05:57 hdbngr has joined #asahi-dev

06:05 zumi_ has joined #asahi-dev

06:06 hdbngr has quit [Ping timeout: 480 seconds]

06:10 zumi has quit [Ping timeout: 480 seconds]

06:10 zumi_ is now known as zumi

06:13 hdbngr has joined #asahi-dev

06:19 lockheed_marten has joined #asahi-dev

06:21 hdbngr has quit [Ping timeout: 480 seconds]

06:28 hightower3 has joined #asahi-dev

06:29 hdbngr has joined #asahi-dev

06:35 hightower2 has quit [Ping timeout: 480 seconds]

06:38 hdbngr has quit [Ping timeout: 480 seconds]

06:46 hdbngr has joined #asahi-dev

06:54 hdbngr has quit [Ping timeout: 480 seconds]

07:02 hdbngr has joined #asahi-dev

07:03 chadmed_ has joined #asahi-dev

07:03 chadmed has quit [Read error: Connection reset by peer]

07:10 hdbngr has quit [Ping timeout: 480 seconds]

07:14 hdbngr has joined #asahi-dev

07:22 hdbngr has quit [Ping timeout: 480 seconds]

07:31 hdbngr has joined #asahi-dev

07:39 hdbngr has quit [Ping timeout: 480 seconds]

07:40 hdbngr has joined #asahi-dev

07:50 hdbngr has quit [Ping timeout: 480 seconds]

07:57 hdbngr has joined #asahi-dev

08:06 hdbngr has quit [Ping timeout: 480 seconds]

08:15 hdbngr has joined #asahi-dev

08:23 hdbngr has quit [Ping timeout: 480 seconds]

08:32 hdbngr has joined #asahi-dev

08:35 <chaos_princess> axi2af bridge tunables should survive a power gate, right?

08:40 hdbngr has quit [Ping timeout: 480 seconds]

08:47 hdbngr has joined #asahi-dev

09:22 Gnappo85 has joined #asahi-dev

09:36 bgb has quit [Ping timeout: 480 seconds]

09:36 roxfan has quit [Ping timeout: 480 seconds]

09:43 Gnappo85 has quit [Quit: Textual IRC Client: www.textualapp.com]

09:50 bgb has joined #asahi-dev

10:16 <jannau> shaved off at least 56 commits of bits/000-devicetree (189 -> 133, latter number includes a couple of queued t2 and iphone changes)

10:18 <jannau> mostly squashing (the speaker enablement history isn't useful) but we also had a few redundant commits

10:22 cylm has quit [Ping timeout: 480 seconds]

10:43 <jannau> do we want to leave ps_atc*_usb_aon always-on? shouldn't be necessary anymore with atcphy

10:43 <jannau> https://github.com/AsahiLinux/linux/commit/ab020063f0b5319c5475df5b92742a975d94a563

11:01 <sven> chaos_princess: never checked, I only know that neither fuses nor the core atcphy tunable survive that (or even a atcphy reset)

11:01 <sven> jannau: I don’t think that necessary anymore, should be easy to test

11:01 <chaos_princess> we set gpu axi2af in m1n1, and i do not think gpu never gets power gated, right?

11:02 <sven> maybe it also just happens to work :D

11:02 <jannau> isn't your for-upstream dwc3/atc-phy branch evidence enough that it's not strictly needed?

11:03 <sven> but I’ll check this weekend for the atcphy axi2af, we could drop the entire axi2af mmio range then. I only know that macos always applies them

11:03 <sven> jannau: oh, yes, I didn’t realize that commit isn’t even upstream

11:05 <jannau> we might have upstreamed it accidentally for t8112 (and my current t602x DT contains it as well)

11:07 <jannau> even upstream t600x has it

11:07 <sven> 🙃

11:07 <sven> we can probably drop that once atcphy is in then

11:09 <jannau> any opinions what to do with the generic compatibles after https://lore.kernel.org/asahi/20250902194528.GA1014943-robh@kernel.org/

11:10 <jannau> I'm splitted with just moving on and using the generic compatibles at least for t602x

11:15 <sven> I don’t care anymore at this point

11:29 roxfan has joined #asahi-dev

11:48 <jannau> chaos_princess: no runtime PM in the gpu driver. firmware does internal power management but that not full power-gating

11:50 <jannau> I guess I'll keep the sent dt-bindings + driver changes to avoid confusion due to the back and forth

11:50 <chaos_princess> ty

12:01 nela4 has quit [Ping timeout: 480 seconds]

12:06 zumi_ has joined #asahi-dev

12:11 zumi has quit [Ping timeout: 480 seconds]

12:11 zumi_ is now known as zumi

13:42 john-cabaj has joined #asahi-dev

13:56 cylm has joined #asahi-dev

13:58 cylm_ has joined #asahi-dev

13:58 cylm has quit [Read error: Connection reset by peer]

16:02 malfunction54 has joined #asahi-dev

16:57 <nickchan-> at this point keeping the dt binding changes are less confusing

16:58 nickchan- has quit [Quit: ZNC 1.9.1+deb2+b3 - https://znc.in]

16:58 nickchan has joined #asahi-dev

17:14 flom84 has joined #asahi-dev

17:14 flom84 has quit [Remote host closed the connection]

17:31 cylm_ has quit [Quit: WeeChat 4.7.0]

17:51 nela4 has joined #asahi-dev

17:57 nela4 has quit [Quit: bye!]

18:06 zumi_ has joined #asahi-dev

18:10 zumi has quit [Ping timeout: 480 seconds]

18:11 zumi_ is now known as zumi

18:25 chadmed_ has quit [Quit: Konversation terminated!]

18:25 chadmed_ has joined #asahi-dev

18:36 ddxtanx has quit [Remote host closed the connection]

18:45 WindowPain6 has joined #asahi-dev

18:51 WindowPain has quit [Ping timeout: 480 seconds]

18:51 WindowPain6 is now known as WindowPain

19:49 ohtwoone has quit [Quit: Ping timeout (120 seconds)]

19:49 ohtwoone has joined #asahi-dev

20:08 hdbngr has quit [Ping timeout: 480 seconds]

20:13 bgb has quit [Remote host closed the connection]

20:13 bgb has joined #asahi-dev

20:33 john-cabaj has quit [Ping timeout: 480 seconds]

20:46 hdbngr has joined #asahi-dev

20:50 <jannau> huh, why is brcmfmac using swiotbl? https://paste.debian.net/1394943/

20:52 ddxtanx has joined #asahi-dev

20:55 hdbngr has quit [Ping timeout: 480 seconds]

20:56 <jannau> the iommu group with wifi/bt is type DMA

20:58 <sven> uh.. is the bus marked as untrusted?

20:58 <sven> i don't remember if that forces swiotlb

21:03 <jannau> doesn't external/untrusted just disable bypass?

21:04 <sven> i thought it also forces anything that's not paged alined through swiotlb

21:04 <sven> *aligned

21:04 <sven> e.g. anything that uses dma_map_sg to directly map buffers instead of using the coherent api to allocate buffers

21:05 <sven> but i don't think we marked the internal pci bus as unsafe

21:12 <jannau> not sure how we would do that. there is the obvious "external-facing" which we don't use

21:13 <sven> i think that's it

21:13 <sven> no idea why it would try to use swiotlb then :/

21:15 <jannau> the size is 16384 so it's page aligned

21:15 <sven> weird

21:19 <sven> hm... not sure that's true looking at the code.

21:19 <sven> it's aligning the size to full page in swiotlb i think

21:20 <sven> but i don't understand how it can ever reach that path unless untrusted is true

21:20 <sven> and that can only happen with external-facing afaict

21:22 <jannau> pci code has external_facing and untrusted

21:23 <jannau> I haven't looked where/how untrusted is set

21:24 bgb has quit [Remote host closed the connection]

21:24 <jannau> based on arch_pci_dev_is_removable()

21:24 bgb has joined #asahi-dev

21:25 <jannau> which is x86 only

21:25 <sven> yeah, that’s from acpi I think

21:26 <sven> essentially their external-facing equivalent

21:29 <jannau> it uses swiotlb if the size is not page alligned: https://elixir.bootlin.com/linux/v6.16.3/source/drivers/iommu/dma-iommu.c#L595

21:30 <sven> Not sure I see how

21:32 <sven> that kmalloc thing should always be false for us, shouldn’t it?

21:36 <jannau> it is not or? "return !dma_kmalloc_safe(dev, dir) && !dma_kmalloc_size_aligned(size);"

21:39 <jannau> i.e. if the size is smaller than 128 or the cache line size

21:39 <sven> we should be dma coherent so kmalloc_safe should be true so !true && whatever should always be false

21:44 nela4 has joined #asahi-dev

21:44 <jannau> I'm not sure I just noticed that I added 'dma-coherent;' only to the t60xx device trees. But I was never sure if that's needed at all

21:44 <jannau> as usual I was tripped up by negation and &&

21:45 <sven> :D

21:45 <sven> I always have to go through it step by step to not mess up

21:46 nela4 has quit []

21:47 nela4 has joined #asahi-dev

21:56 nela4 has quit [Quit: bye!]

21:59 hdbngr has joined #asahi-dev

22:03 nela4 has joined #asahi-dev

22:07 hdbngr has quit [Ping timeout: 480 seconds]

22:16 nela4 has quit [Ping timeout: 480 seconds]

22:42 DarkShadow4444 has quit [Read error: Connection reset by peer]

22:42 DarkShadow44 has joined #asahi-dev

22:55 riker77 has quit [Read error: Connection reset by peer]

22:59 riker77 has joined #asahi-dev

23:21 hdbngr has joined #asahi-dev

23:29 riker77 has quit [Read error: Connection reset by peer]

23:30 hdbngr has quit [Ping timeout: 480 seconds]

23:55 riker77 has joined #asahi-dev