| calebtheythem[m] | Wizzup: i saw those listings, I'm worried they have the same issue as mine where the phone works fine but the panel is just broken | 05:52 |
|---|---|---|
| calebtheythem[m] | the notification led is green in the pictures i saw and from what i can tell that only happens when the phone is booted into Android | 05:53 |
| freemangordon | YAY! no more flickering | 08:06 |
| tmlind | freemangordon: nice, look forward to seeing what the fix is :) | 08:15 |
| freemangordon | tmlind: unfortunately it is un userland | 08:15 |
| freemangordon | and is pvr specific | 08:15 |
| freemangordon | so not sure how that can be ported to kernel | 08:16 |
| freemangordon | bu basically, it is calling PVR2DQueryBlitsComplete on the source pixmap BO before requesting page flip | 08:17 |
| freemangordon | maybe we can teach pvr driver to not signal the exclusive fence it attaches to the BO until all write ops are finished | 08:19 |
| freemangordon | but honestly, this is a bit over my knowledge | 08:19 |
| tmlind | heh ok | 08:36 |
| tmlind | hmm so you're thinking about patching drivers/gpu/drm/pvrsgx kernel driver to somehow know until writes are done? is there any way it would even know? | 08:42 |
| freemangordon | pvr driver know | 08:42 |
| tmlind | ok | 08:42 |
| freemangordon | and also it attaches exclusive fence to the BOs it imports | 08:43 |
| tmlind | so pvr driver would know, omapdrm has no clue, right? | 08:43 |
| freemangordon | omapdrm has clue too | 08:43 |
| tmlind | oh | 08:43 |
| freemangordon | but the 'clue' (fence) is misleading ATM | 08:43 |
| tmlind | ok | 08:43 |
| freemangordon | as pvr driver signal that fence too early | 08:44 |
| freemangordon | this is my understanding at least | 08:44 |
| freemangordon | tmlind: have a look at pvr_linux_fence.c | 08:45 |
| freemangordon | glmark2 running in h-d window, glmark score 53 | 08:52 |
| freemangordon | :) | 08:52 |
| tmlind | nice | 09:01 |
| freemangordon | that should increase, when HW composition is in place | 09:02 |
| freemangordon | but for now is good enough I guess | 09:02 |
| tmlind | yeah and nice to get rid of the old hacks | 09:03 |
| freemangordon | mhm | 09:03 |
| freemangordon | tmlind: https://github.com/maemo-leste/xf86-video-omap/commit/12836b9d1b03e5a9de0b89cba727c801f74b4184 | 09:27 |
| freemangordon | ok, lets see if that fixes flickering on n900 too | 09:30 |
| freemangordon | yes, no flicker :) | 09:42 |
| freemangordon | h-d scrolling fps is a bit low (~30), while modesetting/glamor does ~60, but I guess this is dr3 vs dri2 thing | 09:43 |
| freemangordon | Wizzup: is kernel pakcged for n900 too? | 09:44 |
| tmlind | freemangordon: so is it the sgxMapPixmapBo() that fixes the fence issue? not seeing it paired with sgxUnmapPixmapBo().. i guess i don't follow | 10:08 |
| freemangordon | this https://github.com/maemo-leste/xf86-video-omap/commit/12836b9d1b03e5a9de0b89cba727c801f74b4184#diff-cddd4fe88eaa58d501863ca10266249d311f432eea10411f8a0b08d93a020b55R1018 | 10:09 |
| freemangordon | which ends up calling PVR2DQueryBlitsComplete() on the source pixmap BO | 10:10 |
| tmlind | don't you also want to call sgxUnmapPixmapBo() somewhere? | 10:11 |
| freemangordon | only when pixmap is destroyed | 10:12 |
| freemangordon | sgxMapPixmapBo just gets a handle to call pvr2d functions | 10:12 |
| tmlind | ok i guess you're not seeing memory leaks then :) | 10:12 |
| freemangordon | yeah | 10:12 |
| tmlind | no idea what the generic fix would look like | 10:13 |
| freemangordon | I think fixing pvr fence signal time | 10:13 |
| freemangordon | maybe wite a mail to openpvr ML | 10:13 |
| tmlind | heh | 10:14 |
| tmlind | so what does PVR2DQueryBlitsComplete() end up calling in the kernel? | 10:14 |
| freemangordon | sec | 10:14 |
| tmlind | SGX2DQueryBlitsCompleteKM maybe? | 10:14 |
| freemangordon | no | 10:14 |
| freemangordon | well, yeah | 10:14 |
| freemangordon | but, see https://github.com/TexasInstruments/dri3wsegl/blob/master/pvrhelpers.c#L102 | 10:15 |
| freemangordon | it is what happens basically | 10:15 |
| freemangordon | when you do PVRSRVMapFullDmaBuf on BO, you get 2 things: | 10:16 |
| freemangordon | device address (I think this is the entry in SGX MMU) | 10:16 |
| freemangordon | and PVRSRV_CLIENT_SYNC_INFO structure, which I guess is mmap()-ed kernel memory directly updated by the driver (didn;t look at the details) | 10:17 |
| Wizzup | freemangordon: it might need more work the n900 | 10:17 |
| Wizzup | it's probably in the droid4 component of the repos too | 10:18 |
| freemangordon | it seems every time a render task is started, ui32WriteOpsPending is incremented | 10:18 |
| freemangordon | Wizzup: well, ok, but we want same kernel for all omaps, no? | 10:18 |
| tmlind | freemangordon: oh i sort of recall those counters map back to omapdrm | 10:18 |
| Wizzup | freemangordon: yes | 10:18 |
| freemangordon | the stopper was sgx, but now we have it... :) | 10:19 |
| freemangordon | tmlind: really? never seen those | 10:19 |
| freemangordon | maybe some hack? | 10:19 |
| freemangordon | I mean - never seen those in omapdrm | 10:19 |
| Wizzup | freemangordon: yes, agreed | 10:20 |
| freemangordon | so, we only need to find someone to do the job, no? :p | 10:20 |
| Wizzup | just mentioned to component for testing, but it also needs some other changes wrt /boot/ I think | 10:20 |
| Wizzup | yeah I'll work with parazyd | 10:21 |
| freemangordon | great | 10:21 |
| Wizzup | I think: | 10:21 |
| Wizzup | 1. kernel headers pkg | 10:21 |
| Wizzup | 2. dd pkg | 10:21 |
| Wizzup | ddx | 10:21 |
| freemangordon | will try (1) later on | 10:21 |
| Wizzup | 3. test on d4 | 10:21 |
| Wizzup | 4. make generic and test on n900, bionic | 10:21 |
| Wizzup | right | 10:21 |
| Wizzup | I will work on this today | 10:22 |
| tmlind | freemangordon: i think the pvr driver ops counts map to omapdrm counts in the ddk-1.9 kernel, see the old int sync_op() | 10:23 |
| freemangordon | Wizzup: there are 2 more things to be fixed: | 10:24 |
| freemangordon | 1. even on d4 fps is too low, I think video-omap dri2 implementation is kinda lame, it doesn't allow the client to continue rendering until the current buffer is not flipped. or maybe that's how dri2 works, dunno, lets see what uvos has to say about that. | 10:24 |
| freemangordon | 2. tklock window is not being updated | 10:24 |
| freemangordon | tmlind: ah, ddk 1.9 | 10:24 |
| tmlind | freemangordon: yeah i think both omapdrm and pvr kernel driver tweak the same ops counts.. | 10:25 |
| freemangordon | tmlind: but, doing that in 5.15 is not an option, no? | 10:25 |
| freemangordon | so we need an exclusive fence that gets signalled when renering completes | 10:26 |
| tmlind | not sure, i think we may just need to add checking for write_pending to omapdrm to make sure writes have completed? | 10:26 |
| freemangordon | sorry, don;t get that | 10:26 |
| freemangordon | which code in 5.15 do you mean? | 10:26 |
| tmlind | yeah | 10:27 |
| freemangordon | I mean - how omapdrm even knows of such pvr details? | 10:27 |
| tmlind | so in ddk-1.9 kernel patches, to me it seems that struct omap_gem_object is shared between both omapdrm kernel module and pvr kernel module | 10:27 |
| freemangordon | keep in mind omapdrm already supports fences and wait on exclusive if one exists | 10:28 |
| tmlind | both can tweak the usage counts for write_pending and write_complete | 10:28 |
| freemangordon | so definitely it is pvr that has to be fixed | 10:28 |
| freemangordon | IMO | 10:28 |
| freemangordon | anyway, gtg now, meeting, ttyl | 10:28 |
| tmlind | yeh out of here too, ttyl | 10:29 |
| bencoh | freemangordon: no, sorry (regarding dmafences) | 10:45 |
| Wizzup | freemangordon: right @ things to be fixed | 11:03 |
| Wizzup | freemangordon: do you only need the headers you included (hwdefs and include4), or potentially more | 11:10 |
| Wizzup | this should give us all: find ./drivers/gpu/drm/pvrsgx/1.17.4948957/ -name \*.h -type f -o -type l | 11:10 |
| Wizzup | but that will make it go in /usr/src/linux-headers-5.15.2/include/drivers/gpu/drm/... | 11:10 |
| Wizzup | parazyd: what is the reason we don't just carry the kernel patches in our kernel branches as opposed to in the maemo/beowulf* branches? | 11:16 |
| Wizzup | probably the same question for the 0001-maemo-leste-quirks.patch | 11:16 |
| Wizzup | it is extremely tedious to change anything, since you need to jump back and forth between branches and every time you do there is manual cleanup to be done | 11:16 |
| Wizzup | also the patches are essentially unmaintainable | 11:16 |
| * bencoh nods vigorously | 11:17 | |
| Wizzup | I don't see why we do it this way, the only thing that it might make easier is rebasing on a new branch, but 'git rebase' should just take care of the problems really | 11:18 |
| Wizzup | omg the patches in debian/series aren't even git-am able :( | 11:25 |
| Wizzup | whoever invented quilt must have done it before git | 11:25 |
| Wizzup | freemangordon: oh yeah for n900 we also want the wifi patch | 11:28 |
| Wizzup | parazyd: also what tags should I use to make you and uvos happy | 11:36 |
| Wizzup | any suggestions for the omap device shared kernel | 11:40 |
| Wizzup | (or perhaps even for all devices?) | 11:41 |
| Wizzup | we could do maemo-kernel or omap-kernel | 11:41 |
| Wizzup | freemangordon: with my current (not yet in repo) code it ends up here: | 11:54 |
| Wizzup | /usr/src/linux-headers-5.15.2-218687-g24a714e51fc5-dirty/drivers/gpu/drm/pvrsgx/1.17.4948957/eurasia_km/services4/srvkm/hwdefs/sgx530defs.h | 11:54 |
| Wizzup | ofc. without the -dirty probably | 11:54 |
| Wizzup | all of this is much more messy than just the sgx-ddk-um package I have but hey :P | 11:55 |
| Wizzup | bencoh: latest kernel should be dpkg-buildpackage -b -uc buildable mostly, just checkout maemo-5.15 and then do parazyd's trick to checkout the debian dir | 12:32 |
| Wizzup | I moved most of the patches (apart from config changes) to git | 12:32 |
| bencoh | neat | 12:39 |
| Wizzup | yeah maybe we just disable quilt alltogether later on | 12:40 |
| Wizzup | parazyd: so any suggestions for the name? either maemo-kernel (if we plan on using it for pine64 as well) or omap-kernel | 12:55 |
| Wizzup | or omap-linux / maemo-linux I suppose | 12:55 |
| parazyd | The pine64 kernel is specific to Pinephone/Pinetab, it's not in general for Allwinner. | 12:56 |
| parazyd | So I'd keep that as "pine64-linux". | 12:56 |
| parazyd | For the PVR kernel, I would go with "omap-linux" | 12:56 |
| parazyd | And finally if we want other generic kernels, just "maemo-linux", yeah. | 12:56 |
| parazyd | Or just "linux" even | 12:56 |
| Wizzup | so I am wondering if we shouldn't aim to have omap-linux be the generic one | 12:57 |
| Wizzup | we'll probably always carry some patches on top, but they should not affect other parts | 12:57 |
| parazyd | I think the issue is we build a package with a specific name from the repo. | 12:58 |
| parazyd | And also a specific arch | 12:58 |
| Wizzup | yes, that's the next thing I wanted to bring up | 12:58 |
| parazyd | Not sure how we can have more than 1 | 12:58 |
| Wizzup | what do you mean exactly? | 13:01 |
| parazyd | Building for all these different platforms with a single debian dir | 13:02 |
| Wizzup | right, we'd have to consolidate the package names | 13:02 |
| Wizzup | as in we probably want to get rid of droid4-linux anyway | 13:03 |
| Wizzup | since it's also bionic, droid3, and more | 13:03 |
| Wizzup | and if it's also n900, maybe omap | 13:03 |
| parazyd | Yeah we wanted to rename that to omap-linux | 13:03 |
| Wizzup | but then if we can just use the 5.15 for the pine and others as well, we could just make it maemo-linux | 13:03 |
| Wizzup | was mostly my point | 13:03 |
| Wizzup | I doubt the omap and allwinner patches conflict typically | 13:03 |
| bencoh | you can't really build on the fact that you'll always be able to run the same version on your devices though | 13:04 |
| Wizzup | ok, that argues for omap-linux I guess | 13:04 |
| bencoh | (assuming this fact really plays a part in your decision) | 13:04 |
| parazyd | We can have different branches in the same repo | 13:05 |
| parazyd | But that's not much more beneficial than separate repos | 13:05 |
| Wizzup | saves some bits on hard disks and cloning, but I think it complicates our maemo/* branches | 13:05 |
| Wizzup | no? | 13:05 |
| parazyd | It does | 13:06 |
| bencoh | this will prolly add another layer yeah | 13:06 |
| parazyd | I'll think about it a bit | 13:06 |
| Wizzup | maybe we go for omap-linux for now | 13:06 |
| Wizzup | unrelated, we're getting a HoneyComb LX2 for arm builds | 13:12 |
| freemangordon | Wizzup: do we have pkgconfig for that? | 13:18 |
| freemangordon | (kernel headers) | 13:19 |
| Wizzup | freemangordon: no, not for the new ones, I don't know how to do it for specific kernel versions | 13:20 |
| Wizzup | I had one in sgx-ddk-um but I don't know how to adapt it to path of installed kernel | 13:20 |
| freemangordon | ok, but kernel-headers-dev should have .pc file, no? | 13:20 |
| Wizzup | I don't know of any kernel-headers-dev pkg | 13:21 |
| freemangordon | sorry, kernel-headers or whatever is the name of the package with the headers | 13:21 |
| parazyd | Are you building a kernel module or something else? | 13:21 |
| freemangordon | no, DDX | 13:22 |
| Wizzup | parazyd: no he wants to use the headers in the DDX | 13:22 |
| Wizzup | freemangordon: and no it does not ship with a pc, this is very uncommon I think | 13:22 |
| Wizzup | that's why I copied the headers to sgx-ddk-um | 13:22 |
| parazyd | There's no pkg-config for that | 13:22 |
| freemangordon | can we have sgx-ddk-um depending on a particular kernel-headers version? | 13:23 |
| parazyd | (btw it should be called linux-libc-dev) | 13:23 |
| parazyd | freemangordon: Yes | 13:23 |
| Wizzup | I am against the idea tbh | 13:23 |
| freemangordon | and providing .pc file too | 13:23 |
| Wizzup | I would much rather have sgx-ddm-um use kernel version as versioning | 13:23 |
| Wizzup | and just copy the headers over | 13:23 |
| freemangordon | Wizzup: why? | 13:23 |
| Wizzup | since we're exposing internal kernel driver headers | 13:23 |
| Wizzup | this is never done | 13:23 |
| parazyd | You don't need a .pc file | 13:23 |
| freemangordon | no, those are not internal | 13:23 |
| Wizzup | if you look at normal linux headers they are only for drivers | 13:24 |
| parazyd | They're always in /usr/include/linux | 13:24 |
| Wizzup | DKMS stuff | 13:24 |
| freemangordon | ok, but there are userland interfaces too | 13:24 |
| freemangordon | where are those installed? | 13:24 |
| Wizzup | parazyd: doesn't look like it to me | 13:24 |
| freemangordon | /usr/include, no? | 13:24 |
| Wizzup | merlijn@maemo-leste-armhf:/mnt/merlijn/maemo-leste/droid4-linux$ find /usr/include/linux/ | grep sgx | 13:24 |
| Wizzup | merlijn@maemo-leste-armhf:/mnt/merlijn/maemo-leste/droid4-linux$ dpkg -L linux-headers-droid4 | grep sgx -c | 13:24 |
| Wizzup | 166 | 13:24 |
| Wizzup | the linux header packages are *only* for DKMS | 13:25 |
| freemangordon | the headers ion question are shared between kernel and userland | 13:25 |
| Wizzup | again, please consider just having the headers in sgx-ddm-um pkg | 13:25 |
| freemangordon | Wizzup: I understand your point, but we'll have conflict if one tries to install both packages | 13:25 |
| Wizzup | we can pin/match them exactly | 13:26 |
| freemangordon | doesn't matter, dpkg will complain | 13:26 |
| Wizzup | without having convoluted paths for the headers and without exporting kernel specific headers | 13:26 |
| Wizzup | huh? | 13:26 |
| freemangordon | "...tries to overwrite a file which is also in ..." | 13:26 |
| Wizzup | no | 13:26 |
| freemangordon | and having same files in two different places is not a good idea imo | 13:27 |
| Wizzup | the only reason the pvrsgx headers are currently exported at all in any package is because I hacked it in for you | 13:27 |
| Wizzup | they are otherwise never in any headers package since as far as kernel is considered they are internal | 13:27 |
| Wizzup | they are not part of any public api/abi | 13:27 |
| freemangordon | also, keep in mind that because of iphb dkms module we'll have kernel headers installed either way | 13:27 |
| Wizzup | so there will be no conflict if we have them in sgx-ddk-um | 13:27 |
| Wizzup | that is unrelated | 13:27 |
| Wizzup | just dpkg -L linux-headers-droid4 now and see how little it is | 13:28 |
| freemangordon | Wizzup: we'll need sgx-ddm-um either ways though | 13:28 |
| freemangordon | because of pvr2d.h | 13:28 |
| Wizzup | yes, that's already packaged | 13:28 |
| Wizzup | with the headers currently | 13:28 |
| Wizzup | but then you told me to use the kernel package instead | 13:28 |
| Wizzup | which is fine, but I don't know how to make a .pc file for it | 13:29 |
| freemangordon | do not | 13:29 |
| freemangordon | put .pc file in sgx-ddm-um | 13:29 |
| Wizzup | it is | 13:29 |
| Wizzup | uh | 13:29 |
| freemangordon | and make it depend on "kernel-headers" package | 13:29 |
| freemangordon | the particular version that is | 13:30 |
| Wizzup | did you see how ugly the paths are? | 13:30 |
| Wizzup | /usr/src/linux-headers-5.15.2-218687-g24a714e51fc5-dirty/drivers/gpu/drm/pvrsgx/1.17.4948957/eurasia_km/include4/pvr_debug.h | 13:30 |
| freemangordon | yeah | 13:30 |
| Wizzup | how do I make a pc file fro that? | 13:30 |
| freemangordon | hmm | 13:30 |
| Wizzup | also, *every* other sgx release we did we had the headers in a separate repo | 13:30 |
| Wizzup | afaik | 13:30 |
| freemangordon | oh, ok :) | 13:30 |
| freemangordon | do it as you think is best. | 13:31 |
| Wizzup | if you have an idea on how we can make the .pc file we can make it work | 13:31 |
| Wizzup | otherwise I would rather have the header files in the sgx-ddk-um pkg | 13:31 |
| Wizzup | but afaik pkgconfig cannot script or use globs in paths | 13:32 |
| freemangordon | what about creating sgx-ddk-um-headers pkg created out of kernel build? | 13:32 |
| Wizzup | (so /usr/src/linux-headers-*/driver/gpu/drm/pvrsgx/... is out of the question | 13:32 |
| Wizzup | ) | 13:32 |
| freemangordon | you can create package.pc.in | 13:32 |
| Wizzup | yes, that is potentially possible, but it's quite some work because kernel ships with it's own builddeb script | 13:32 |
| freemangordon | ah | 13:32 |
| Wizzup | yeah... :p | 13:32 |
| Wizzup | I haven't really seen .pc files be shipped from any kernel pkg much either | 13:33 |
| freemangordon | oh, ok then | 13:33 |
| freemangordon | just make sure sgx-ddk-um-dev provides "sgx-ddk-um-1.17.4948957" or somesuch | 13:33 |
| Wizzup | and regarding versioning, shall we postfix it with kernel version? | 13:34 |
| Wizzup | or maybe prefix | 13:34 |
| freemangordon | I don;t think kernel version matters | 13:34 |
| freemangordon | I think it is DDK version that does | 13:34 |
| Wizzup | it will if you end up changing headers | 13:34 |
| Wizzup | as in I was assuming you planned to change the headers | 13:35 |
| freemangordon | no | 13:35 |
| Wizzup | oh | 13:35 |
| freemangordon | what to change there? | 13:35 |
| Wizzup | well, I just assumed you would change them because you wanted them from the kernel | 13:35 |
| Wizzup | *shrug* | 13:35 |
| Wizzup | let me remove the pvr headers from the next kernel build again, and I'll try to tidy up the sgx-ddk-um as you request | 13:36 |
| freemangordon | ok | 13:36 |
| Wizzup | I was already able to build the DDX with it earlier | 13:36 |
| Wizzup | the kernel that is building atm should also work on n900 (although it might not ship with the dtb - we need to fix that) | 13:36 |
| freemangordon | the point is that DDX should depend on sgx -dev package providing headers for a particular DDK version | 13:37 |
| Wizzup | anything else, otherwise I'll take a quick break? | 13:37 |
| Wizzup | ok | 13:37 |
| Wizzup | makes sense | 13:37 |
| freemangordon | but I'll leave that to you and parazyd to agree upon how to do | 13:37 |
| Wizzup | I'll make sure to do that | 13:37 |
| freemangordon | ok | 13:37 |
| Wizzup | I plan to have my droid4 and n900 over to 5.15 by this evening :) | 13:37 |
| Wizzup | at least in experimental | 13:37 |
| freemangordon | Wizzup: btw, did you see that I made n900 flicker go away? | 13:37 |
| Wizzup | I read it yes :-D | 13:38 |
| freemangordon | I need uvos though, I want to discuss with him if it makes sense to implement dri3 | 13:38 |
| Wizzup | ok | 13:38 |
| freemangordon | or to fix dri2 | 13:38 |
| Wizzup | but I think we can move forward with this, at least for experimental | 13:38 |
| Wizzup | maybe even -devel | 13:39 |
| freemangordon | yeah | 13:39 |
| freemangordon | mind you, still no portrait on n900 | 13:39 |
| freemangordon | and it is slower that currently | 13:39 |
| freemangordon | *than | 13:39 |
| Wizzup | *shrug* it's better than 5.1 | 13:39 |
| Wizzup | I use the n900 mostly for devel/testing, so I don't mind that so much | 13:39 |
| freemangordon | not sure its better, but it will be :) | 13:39 |
| Wizzup | I think it's better to do the kernel merging now | 13:40 |
| freemangordon | sure | 13:40 |
| Wizzup | it also allows us to toy with OFF mode | 13:40 |
| crab | hi all, i updated my maemo-leste the other day and the gui appears to have died a bit, maybe because it wants ofono? its not a big issue as i mainly use it headless for irc (its lurking in this channel as n900) but if anyone has any suggestions id be grateful. :) | 13:40 |
| sicelo | +1 @ OFF mode. | 13:40 |
| Wizzup | crab: yikes you're the third to report that | 13:40 |
| crab | im just trying another update now so apologies in advance if that fixes it. :) | 13:40 |
| Wizzup | brb sorry, half an hour or so | 13:41 |
| crab | Wizzup: is that bad or good? | 13:41 |
| crab | my installation is a bit crazy so im reluctant to moan too much as it could well be stuff like that! | 13:41 |
| crab | also <smugly>sorry for registering 'n900' for my n900 if anyone wanted to do that, but it had to be done!</smugly> | 13:42 |
| sicelo | btw, although i can't say 100% certainty atm, with 5.15 expect other N900 issues, e.g. non-working RGB LED, etc. But I know why they're like that ... just didn't have time to make proper fixes and send upstream | 13:42 |
| Wizzup | sicelo: what are your fixes? | 13:42 |
| Wizzup | we could include them now | 13:42 |
| Wizzup | crab: :p | 13:42 |
| Wizzup | actually brb now :) | 13:42 |
| sicelo | turned out to not be a good fix, imo. so the rgb led controller doesn't get turned on at boot or when you probe the module. so i had proposed to have it set on at boot. LED subsystem maintainers (pavel) were willing to accept it. | 13:45 |
| sicelo | it's turned on by GPIO. I wanted to set it always high. later on, someone had a patch where he needed to set it always low :-p | 13:46 |
| sicelo | so yes, that's why i think it needs further tracking | 13:46 |
| sicelo | https://www.spinics.net/lists/linux-leds/msg18183.html , https://www.spinics.net/lists/linux-leds/msg18195.html , https://www.spinics.net/lists/linux-leds/msg19029.html | 13:47 |
| sicelo | atm i don't remember the other guy's commit, but he was editing this exact line on the driver | 13:48 |
| sicelo | so you could take my fix (always set gpio high), with the understanding that a better solution must be found | 13:53 |
| sicelo | anyway also, as said earlier, i haven't played with recent kernels. maybe something has changed already with these drivers and maybe will work correctly ootb now | 13:54 |
| Wizzup | parazyd: we also need to ship/install more dtbs I think | 14:27 |
| Wizzup | parazyd: what section should we use? | 14:36 |
| Wizzup | omap? | 14:36 |
| Wizzup | (and how do we have people upgrade to that) | 14:36 |
| Wizzup | e.g. xf86-video-omap is section droid4 | 14:37 |
| Wizzup | freemangordon: what version do you want for the ddx in the changelog | 14:43 |
| Wizzup | 0.4.6 or 0.5.0 | 14:43 |
| Wizzup | I am assuming 0.5.0 | 14:44 |
| freemangordon | yeah | 14:50 |
| Wizzup | before the end of the year we will have a much faster build machine | 14:53 |
| Wizzup | for arm | 14:53 |
| Wizzup | freemangordon: ok, should be in the repo in ~2 mins | 14:54 |
| Wizzup | I guess that's all the components | 14:54 |
| Wizzup | any hint on xorg.conf required? | 14:56 |
| Wizzup | I could try it now on my droid4 | 14:56 |
| freemangordon | keep in mind this will still not work on n900 | 15:01 |
| freemangordon | but on d4 should be ok | 15:01 |
| Wizzup | ok | 15:02 |
| Wizzup | I pushed to xf86-video-omap | 15:02 |
| Wizzup | I think we might need replacements for some of these: pvr-omap4 pvr-omap4-libs pvr-omap4-utils | 15:16 |
| Wizzup | or, not sure you did you the upgrade | 15:16 |
| Wizzup | can I tell to remove certain packages and install other ones or something? | 15:17 |
| Wizzup | does our mesa not provide libgles1 ? | 15:19 |
| Wizzup | heh hildon-meta-droid4 wants pvr-omap4 | 15:21 |
| Wizzup | and cloudgps hard depends on it it seems like | 15:21 |
| * Wizzup documents | 15:21 | |
| Wizzup | freemangordon: so any xorg.conf stuff I should change, or anything else, before I reboot? | 15:23 |
| Wizzup | e.g. to enable exa? | 15:23 |
| Wizzup | https://dpaste.com/GX369J8NC | 15:25 |
| Wizzup | rebooting | 15:25 |
| Wizzup | let's see | 15:26 |
| Wizzup | well it booted to h-d on first go | 15:28 |
| Wizzup | it doesn't use EXA I think | 15:28 |
| Wizzup | [ 33.217] (WW) Warning, couldn't open module omap_pvr | 15:28 |
| Wizzup | [ 33.217] (EE) OMAP: Failed to load module "omap_pvr" (module does not exist, 0) | 15:28 |
| Wizzup | [ 33.218] (II) OMAP(0): Cannot load the omap_pvr sub-module | 15:28 |
| Wizzup | [ 33.226] (II) OMAP(0): Soft EXA mode | 15:28 |
| Wizzup | no more pvr panics in kernel is *nice* | 15:30 |
| Wizzup | s/panics/oopses/ | 15:30 |
| Wizzup | I guess it needs the mesa env change or something | 15:32 |
| sicelo | iirc exa is used (only?) when scrolling | 15:37 |
| Wizzup | so /etc/profile.d/mesa.sh is not read when X is starting | 15:38 |
| Wizzup | # cat /proc/`pidof Xorg`/environ | grep -a MESA | 15:38 |
| Wizzup | I wonder where fmg defines this env var | 15:38 |
| Wizzup | I'm a bit stuck right now until I know what the alternative fix is, symlink? | 15:40 |
| Wizzup | also 5.15 has some problems initialising wifi some of times it looks like, but that's for later.. | 15:42 |
| Wizzup | freemangordon: I did this: | 15:44 |
| Wizzup | /usr/lib/arm-linux-gnueabihf/dri/ | 15:44 |
| Wizzup | ln -s pvr_dri.so omap_dri.so | 15:44 |
| Wizzup | but that still leaves the above error | 15:44 |
| Wizzup | maybe my symlink should be not relative | 15:45 |
| Wizzup | I guess I'll take a break :) | 15:46 |
| Wizzup | maybe it's not packaged or something | 15:48 |
| Wizzup | yeah it's not in the package.. | 15:49 |
| Wizzup | now it hangs at: | 15:58 |
| Wizzup | [ 123.652] (II) OMAP(0): Initializing the "omap_pvr" sub-module ... | 15:58 |
| * Wizzup giving up for now | 15:58 | |
| Wizzup | ah: /usr/lib/xorg/Xorg: symbol lookup error: /usr/lib/xorg/modules/omap_pvr_drv.so: undefined symbol: PVRSRVConnect | 15:59 |
| Wizzup | freemangordon: what is your omap_pvr_drv.so linked against? | 16:01 |
| Wizzup | oh.. | 16:07 |
| Wizzup | ok, that is solved, but now it cannot enumerate pvr devices | 16:10 |
| Wizzup | so it is not initialised | 16:10 |
| Wizzup | iirc you told me not to package that, freemangordon ? | 16:10 |
| Wizzup | enough of this monologue, ttyl :) | 16:11 |
| Wizzup | looks like it needs pvrsrvctl for sure | 16:13 |
| Wizzup | freemangordon: now I get this: | 16:18 |
| Wizzup | PVR:(Error): LoadCompilerModule: Couldn't load library libglslcompiler.so [0, ] | 16:18 |
| Wizzup | and ldd gives: | 16:19 |
| Wizzup | # ldd /usr/lib/arm-linux-gnueabihf/libglslcompiler.so | 16:19 |
| Wizzup | /usr/lib/arm-linux-gnueabihf/libglslcompiler.so: /lib/arm-linux-gnueabihf/libm.so.6: version `GLIBC_2.29' not found (required by /usr/lib/arm-linux-gnueabihf/libglslcompiler.so) | 16:19 |
| Wizzup | so I think this still needs the libc treatment | 16:19 |
| Wizzup | I forgot where I had the code to soften those requirements | 16:21 |
| * Wizzup break for real | 16:21 | |
| freemangordon | Wizzup: sorry, I was on a meeting | 17:07 |
| freemangordon | yes, I will push a couple of fixes, one of them being the init fix | 17:07 |
| freemangordon | and yes, as uvos said we have to relax libglslcompiler.so glibc requirement | 17:08 |
| freemangordon | script to do that is on github issue, LMK if you cannot find it, I have the script around | 17:08 |
| freemangordon | Wizzup: which branch shall I continue work? | 17:13 |
| freemangordon | Wizzup: sgx-ddk-um-dev should depend on sgx-ddk-um, no? | 17:16 |
| freemangordon | Wizzup: did you push the solution to "undefined symbol: PVRSRVConnect"? | 17:24 |
| freemangordon | Wizzup: going on a walk, ttyl, please push if you gave fix for ^^^ | 17:25 |
| Wizzup | freemangordon: hi | 17:29 |
| Wizzup | freemangordon: I need libc fixes only | 17:29 |
| Wizzup | freemangordon: problem with that symbol is stray + in my Makefile.am | 17:29 |
| Wizzup | I have fix here | 17:30 |
| freemangordon | please push it | 17:30 |
| Wizzup | it is on maemo/beowulf-experimental | 17:30 |
| freemangordon | is not, just pulled | 17:30 |
| Wizzup | oh it is not\ | 17:30 |
| Wizzup | sry just got back | 17:30 |
| Wizzup | it is now | 17:31 |
| freemangordon | np, take your time | 17:31 |
| freemangordon | ok | 17:31 |
| Wizzup | nah I want to see h-d on my device so I'm happy to prioritise :p | 17:31 |
| freemangordon | Wizzup: https://pastebin.com/Z83FPY8u | 17:31 |
| Wizzup | ok | 17:31 |
| Wizzup | ty | 17:31 |
| freemangordon | will take few more hors, I need to go out for a while | 17:31 |
| Wizzup | the init thing? | 17:31 |
| Wizzup | or something else? | 17:31 |
| freemangordon | I've been home all day long, need some fresh air | 17:32 |
| freemangordon | yes, the init thing | 17:32 |
| freemangordon | 1-2 hours | 17:32 |
| freemangordon | and 2 more minor fixes | 17:32 |
| freemangordon | anyway, ttyl | 17:32 |
| Wizzup | ttyl | 17:34 |
| Wizzup | I have to wait for my d4 battery to discharge anyway, it's too full so usbnet fails 100% | 17:34 |
| Wizzup | works | 17:37 |
| Wizzup | cool | 17:37 |
| Wizzup | maep is smooth :-) | 17:38 |
| Wizzup | "wow" | 17:38 |
| Wizzup | 5.15 doesn't always boot it looks like to me | 17:46 |
| _uvos_ | Wizzup: i added the script on the ddk1.17 bug | 19:16 |
| _uvos_ | oh im behind | 19:23 |
| Wizzup | :) | 19:24 |
| Wizzup | yeah it's running now | 19:24 |
| Wizzup | I will fix up the packaging after dnner | 19:24 |
| _uvos_ | great | 19:26 |
| _uvos_ | ill see ich i can debug why it wont boot | 19:26 |
| _uvos_ | do you have kernel console on uart? | 19:26 |
| _uvos_ | that has this as known issue | 19:27 |
| Wizzup | I can check console later yeah | 19:27 |
| Wizzup | it is random | 19:27 |
| _uvos_ | having the console on uart _causes_ this issue | 19:28 |
| _uvos_ | it is one cause at least | 19:28 |
| _uvos_ | thats why its disabled | 19:28 |
| _uvos_ | in the default bootentry atm | 19:28 |
| Wizzup | aha | 19:28 |
| _uvos_ | there is also a uart interrupt throtelling patch | 19:29 |
| _uvos_ | to help here | 19:29 |
| _uvos_ | it might have gohn missing | 19:29 |
| _uvos_ | ill look when im @home | 19:29 |
| _uvos_ | 1h approx | 19:29 |
| freemangordon | Wizzup: https://github.com/maemo-leste/xf86-video-omap/commit/78249f8721b631d05ba516d4ca8bb4bfc0cdf5aa | 19:41 |
| freemangordon | one more patch and we are ready | 19:42 |
| freemangordon | but first -dinner | 19:42 |
| parazyd | Unpacking sgx-ddk-um-ti443x (1.17.4948957-leste1+2m7.1) ... | 19:57 |
| parazyd | dpkg: error processing archive /var/tmp/apt-dpkg-install-vPgEQ7/12-sgx-ddk-um-ti443x_1.17.4948957-leste1+2m7.1_armhf.deb (--unpack): | 19:57 |
| parazyd | trying to overwrite '/usr/lib/arm-linux-gnueabihf/libPVRScopeServices.so', which is also in package pvr-omap4-libs 1.9.0.8.1.3-1+2m7.4 | 19:58 |
| parazyd | I think here we need a Conflicts, Replaces, Provides | 19:58 |
| parazyd | freemangordon, Wizzup: agreed? ^ | 19:58 |
| parazyd | That's https://github.com/maemo-leste/pvr-omap4 | 20:01 |
| Wizzup | freemangordon: cool, I'll build a new one momentarily | 20:14 |
| Wizzup | parazyd: hmm what? | 20:15 |
| Wizzup | parazyd: yes we need the replaces for the pvr-omap4 stuff in sgx-ddk-um | 20:15 |
| Wizzup | another thing is that we want the rights libs to be picked for n900 | 20:15 |
| Wizzup | so it needs to depend on those specifically somehow | 20:15 |
| Wizzup | (same for droid4 really, we just got lucky with what gets picked) | 20:15 |
| Wizzup | parazyd: but the commands I sent to you remove the packages first | 20:16 |
| Wizzup | freemangordon: check | 20:16 |
| Wizzup | freemangordon: so the ones you added for command mode displays, is that for n900? | 20:20 |
| Wizzup | or does n900 need more to render in landscape | 20:20 |
| Wizzup | freemangordon: building new 0.5.0 version | 20:21 |
| freemangordon | Wizzup: yes, that one patch is for n900 | 20:23 |
| freemangordon | lemme push it | 20:23 |
| Wizzup | ah ok | 20:23 |
| Wizzup | is it not already in maemo/beowulf-experimental? | 20:23 |
| Wizzup | I'll just do anoter build np | 20:23 |
| Wizzup | should we weaken the symbols postinst or shall we just do it in the package itself? | 20:24 |
| Wizzup | i.e. I'll weaken them in git | 20:24 |
| freemangordon | better weaken in git if we keep a cpoy | 20:25 |
| freemangordon | *copy | 20:25 |
| Wizzup | agreed | 20:25 |
| freemangordon | pushed | 20:26 |
| uvos | Wizzup: btw | 20:27 |
| uvos | Wizzup: havent tried this but dosent disabeling charging make usbnet work? as a workaround | 20:27 |
| Wizzup | freemangordon: ok will do another build in a bit | 20:28 |
| Wizzup | uvos: yes but that doesn't work if I have no access | 20:28 |
| Wizzup | :p | 20:28 |
| uvos | emergency shell? | 20:29 |
| Wizzup | freemangordon: btw I force pushed to xf86-video-omap after I did a rebase and also re-created tags | 20:29 |
| Wizzup | uvos: sure I could have done that but it was better to wait 1 minute | 20:29 |
| Wizzup | faster anyway | 20:29 |
| uvos | ok | 20:29 |
| freemangordon | Wizzup: I did a rebase and push | 20:30 |
| Wizzup | freemangordon: ok cool | 20:30 |
| freemangordon | Wizzup: how to install the kernel? | 20:31 |
| Wizzup | freemangordon: let me paste what I did | 20:32 |
| uvos | what are we missing for final drop to devel? | 20:32 |
| uvos | need any help on anything? | 20:32 |
| freemangordon | uvos: lets have it for a while in experimental first :) | 20:32 |
| Wizzup | https://pastebin.com/raw/sYazkHC8 | 20:32 |
| Wizzup | I did that^ | 20:32 |
| Wizzup | uvos: yeah let's give in ~1-3 days, what's needed is to make the transition smooth and also make it work on n900 etc | 20:33 |
| Wizzup | omap-linux is one of those, and there's more to do | 20:33 |
| uvos | freemangordon: fine by me | 20:33 |
| uvos | second question still stands :) | 20:33 |
| freemangordon | ah, omap-linux | 20:33 |
| freemangordon | uvos: yeah, you may want to look why dri2 does single-buffer | 20:33 |
| freemangordon | like, new buffer is requested only after SwapComplete is generated | 20:34 |
| uvos | ok - in what context, it should do single buffer no | 20:34 |
| freemangordon | why not double-buffering? | 20:34 |
| Wizzup | freemangordon: oh we need to deal with that symlink thing | 20:35 |
| Wizzup | with this: ln -s /usr/lib/arm-linux-gnueabihf/dri/pvr_dri.so /usr/lib/arm-linux-gnueabihf/dri/omap_dri.so | 20:35 |
| freemangordon | no, whay | 20:35 |
| freemangordon | *why | 20:35 |
| Wizzup | I set the env var in /etc/profile.d/mesa.sh | 20:35 |
| Wizzup | and it doesn't work | 20:35 |
| Wizzup | X doesn't pick it up | 20:35 |
| uvos | well x is historily just not double bufferd, unless you do composing | 20:35 |
| freemangordon | Wizzup: sure it does | 20:35 |
| Wizzup | are you _sure_? | 20:35 |
| freemangordon | uvos: well, I guess h-d does compositing :) | 20:35 |
| freemangordon | Wizzup: yes | 20:36 |
| Wizzup | # cat /etc/profile.d/mesa.sh | 20:36 |
| Wizzup | export MESA_LOADER_DRIVER_OVERRIDE=pvr | 20:36 |
| Wizzup | I had this | 20:36 |
| Wizzup | rebooted | 20:36 |
| Wizzup | and then grepped /proc/`pidof Xorg`/environ | 20:36 |
| Wizzup | and it *was not* there | 20:36 |
| Wizzup | same for dsme | 20:36 |
| freemangordon | shebang is wrong | 20:36 |
| Wizzup | oh | 20:36 |
| freemangordon | root@devuan-droid4:~/xorg/xf86-video-omap# cat /etc/profile.d/mesa.sh | 20:36 |
| freemangordon | #!/bin/sh | 20:36 |
| freemangordon | export MESA_LOADER_DRIVER_OVERRIDE=pvr | 20:36 |
| freemangordon | :) | 20:36 |
| Wizzup | hmm but it was there for osso-xterm | 20:36 |
| Wizzup | ok, and then who provides that script? | 20:37 |
| freemangordon | um-ddk-libs? | 20:37 |
| Wizzup | ok | 20:37 |
| Wizzup | parazyd: ^ ok? | 20:37 |
| uvos | well the compositing manager is supposed to internaly keep the second buffer | 20:37 |
| uvos | some ddx do do double buffering -intel for example | 20:38 |
| uvos | it you want that then sure i could take a look how its implemented there | 20:38 |
| uvos | otherwise i lack information/knowalge on whats wrong | 20:38 |
| freemangordon | uvos: we have 2 buffers, it is just that no new one is requested until page flip happens | 20:38 |
| uvos | ah | 20:39 |
| uvos | ok | 20:39 |
| uvos | yeah | 20:39 |
| freemangordon | which basically means single-buffer | 20:39 |
| uvos | yeah | 20:39 |
| parazyd | Wizzup: Yeah I don't want to use the dpkg commands to remove the packages. Rather make a proper upgrade. | 20:39 |
| uvos | thats not right | 20:39 |
| freemangordon | so h-d renders with 40fps | 20:39 |
| Wizzup | parazyd: check | 20:39 |
| parazyd | Wizzup: If I delete them, then it's more difficult to reproduce the situation | 20:39 |
| freemangordon | uvos: :nod: | 20:39 |
| Wizzup | parazyd: ok, yeah, agreed, fyi I changed sgx-ddk-um a bit so make sure to pull | 20:39 |
| parazyd | *nod* | 20:39 |
| freemangordon | enable debug in omap.conf and you'll see | 20:39 |
| uvos | ok | 20:40 |
| uvos | ill have to get all this stuff on some d4 first | 20:40 |
| Wizzup | in 10 mins I should have latest ddx and sgx-ddm-uk built | 20:41 |
| Wizzup | ddk-um* lol | 20:41 |
| freemangordon | Wizzup: hmm, I see neither omap-linux nor linux-omap | 20:41 |
| freemangordon | droid4-linux? | 20:42 |
| Wizzup | freemangordon: the package is in the paste I sent you | 20:43 |
| Wizzup | apt upgrade should just work though for it | 20:43 |
| Wizzup | freemangordon: yeah ok so it wasn't in the paste I just apt upgraded | 20:43 |
| Wizzup | it's linux-image-droid4 | 20:43 |
| freemangordon | linux-image-droid4/testing 5.11.2-3+2m7 armhf [upgradable from: 5.11.2-2+2m7] | 20:44 |
| Wizzup | apt update? | 20:44 |
| freemangordon | mhm | 20:44 |
| Wizzup | I mean I got 5.15 from the repo so it must be there | 20:44 |
| freemangordon | what is the name according to dlkg -l ? | 20:45 |
| freemangordon | and version | 20:45 |
| Wizzup | ii linux-image-droid4 5.15.0-1+2m7.2 armhf Linux kernel for Motorola Droid4 | 20:45 |
| freemangordon | E: Version '5.15.0-1+2m7.2' for 'linux-image-droid4' was not found | 20:46 |
| Wizzup | well, please double check your repo setup and see if apt update complains | 20:46 |
| freemangordon | deb https://maedevu.maemo.org/leste beowulf-experimental main contrib non-free | 20:46 |
| freemangordon | ah, d4? | 20:47 |
| freemangordon | I was expecting this to hit main :) | 20:47 |
| Wizzup | yes it's in droid4 component atm | 20:47 |
| freemangordon | ok, that explains it | 20:48 |
| Wizzup | are you trying on some other device? | 20:48 |
| freemangordon | no | 20:48 |
| Wizzup | ok, if you try it on n900 it will not work since we don't ship the dts | 20:48 |
| freemangordon | yeah | 20:48 |
| freemangordon | btw, what's with hildon-connectivity-wlan? | 20:49 |
| freemangordon | I cannot upgrade it | 20:49 |
| freemangordon | rebooting, lets see | 20:55 |
| freemangordon | YAY! it's alive :) | 20:56 |
| Wizzup | :) | 20:56 |
| freemangordon | but, with SW rendering :) | 20:56 |
| Wizzup | hm? | 20:57 |
| freemangordon | lemme check | 20:57 |
| Wizzup | sry [hone | 20:58 |
| freemangordon | PVR:(Error): PVRSRVBridgeCall: Failed to access device. Function ID:3223086849 (strerror returns no value.). [0, ] | 20:59 |
| freemangordon | oh, seems x gets initialized before pvr module :( | 21:00 |
| uvos | pvrsrvinit missing? | 21:02 |
| freemangordon | DDX replaces it | 21:03 |
| uvos | no | 21:03 |
| uvos | thats not ok | 21:03 |
| freemangordon | yes, it is | 21:03 |
| freemangordon | sec | 21:03 |
| uvos | its in sysinit for a reason | 21:03 |
| uvos | we need it for sdl usage before x | 21:03 |
| freemangordon | oh | 21:03 |
| uvos | for charging-sdl to work on | 21:03 |
| uvos | drm | 21:03 |
| freemangordon | ok, I see | 21:03 |
| freemangordon | ok, where shall we have pvrsrvinit? | 21:04 |
| uvos | yeah just this https://github.com/IMbackK/pvr-omap4/blob/master/pvrsrvinit.c | 21:05 |
| uvos | is fine | 21:05 |
| freemangordon | yeah, ok, but who shall call it? | 21:05 |
| uvos | openrc | 21:05 |
| uvos | in sysinit | 21:05 |
| Wizzup | I can add te init script | 21:05 |
| Wizzup | we have it already for old ddk | 21:05 |
| freemangordon | yes, please do | 21:05 |
| freemangordon | hmm, something got broken | 21:09 |
| freemangordon | I can't open applications | 21:09 |
| freemangordon | oh, I should have that TS specific udev rule, right? | 21:15 |
| uvos | yeah | 21:15 |
| uvos | ofc | 21:15 |
| uvos | I can't open applications = you cant open the application menu? | 21:15 |
| uvos | yeah | 21:15 |
| uvos | your clicking the hildon button twice :) | 21:16 |
| freemangordon | yeah | 21:16 |
| freemangordon | I think I am going crazy :) | 21:16 |
| freemangordon | *thought | 21:16 |
| freemangordon | Wizzup: run glmark from h-d :p | 21:21 |
| freemangordon | ok, we have issue with fullscreen applications | 21:23 |
| freemangordon | maybe I shall flush framebuffer from exa copy function too | 21:24 |
| Wizzup | freemangordon: back | 21:24 |
| Wizzup | I will port the init script to sgx-ddk-um now | 21:24 |
| freemangordon | ok | 21:24 |
| Wizzup | and add the env | 21:25 |
| Wizzup | (profile.d) | 21:25 |
| freemangordon | I am going to have some rest for a while | 21:25 |
| Wizzup | *nod* | 21:25 |
| freemangordon | we have one issue remaning (fullscreen apps), I'll take care of it maybe tomorrow | 21:25 |
| Wizzup | is that tklock | 21:25 |
| freemangordon | Wizzup: please run glmark in window and tell me what you think | 21:26 |
| Wizzup | I hope to get n900 going by the end of the day | 21:26 |
| Wizzup | ok, is it in our repo? | 21:26 |
| freemangordon | tklock and xterm | 21:26 |
| Wizzup | I always forget where we get it from | 21:26 |
| freemangordon | glmark? | 21:26 |
| freemangordon | no | 21:26 |
| Wizzup | dorian scroling is smoother I think | 21:26 |
| Wizzup | :) | 21:26 |
| freemangordon | I clone/compile | 21:26 |
| Wizzup | and maep is much muhc better | 21:26 |
| Wizzup | ok | 21:26 |
| Wizzup | freemangordon: hm I see some corruption now after reboot | 21:27 |
| Wizzup | could be your flip change? | 21:28 |
| Wizzup | wifi scan dialog has random parts of characters/flyphs | 21:29 |
| Wizzup | glyphs* | 21:29 |
| freemangordon | Wizzup: yeah, saw it too | 21:31 |
| freemangordon | maybe it is https://github.com/maemo-leste/xf86-video-omap/commit/ecd60e09c289db00f7bc83318741099934342fd5 | 21:32 |
| freemangordon | please revert it and try if it fixes the corruption | 21:32 |
| Wizzup | ok | 21:32 |
| freemangordon | I'll make a better patch that skips that wait for fullscreen DRI blits only | 21:33 |
| freemangordon | if it turns out this is the one to blame | 21:34 |
| freemangordon | Wizzup: or, bisect :) | 21:35 |
| freemangordon | sorry about that | 21:35 |
| Wizzup | np | 21:39 |
| Wizzup | I'll start with the obvious target | 21:46 |
| Wizzup | /commit | 21:46 |
| parazyd | Wizzup, freemangordon: Do we still need /usr/bin/pvrsrvinit ? | 21:57 |
| parazyd | (We have that from pvr-omap4-utils | 21:57 |
| parazyd | ) | 21:57 |
| Wizzup | yes | 21:57 |
| Wizzup | I am on it | 21:57 |
| Wizzup | for sgx-ddk-um | 21:57 |
| parazyd | ah ok | 21:57 |
| parazyd | Can you please add the conflicts then too? | 21:57 |
| parazyd | I'll give you a diff | 21:57 |
| Wizzup | hm? | 21:58 |
| parazyd | Are you working on the package right now? | 21:58 |
| parazyd | (If so, just add this as well: http://ix.io/3FO8) | 21:58 |
| Wizzup | phone 1min... | 21:58 |
| uvos | dont use the closed source pvrsrvinit binary ofc | 22:00 |
| uvos | no need | 22:00 |
| Wizzup | parazyd: please add whatever the pkg needs | 22:04 |
| Wizzup | I will just add the init script and the profile.d | 22:04 |
| Wizzup | but pvr-omap4 is deprecated so need to work on that... | 22:04 |
| Wizzup | parazyd: I wasn't working on the package, but I will add some stuff now | 22:15 |
| Wizzup | after I figure out the corruption | 22:16 |
| Wizzup | parazyd: basically we just need the initscript from pvr-omap4-utils in sgx-ddk-um-libs I think | 22:16 |
| Wizzup | you can make it a separate non -libs thing | 22:16 |
| Wizzup | also for testing the n900 with droid4-linux all we need to do is add a dtb file to the builddeb script no? | 22:17 |
| uvos | you have to build uimage too or? | 22:17 |
| uvos | or dose n900 uboot work with zimage now | 22:17 |
| Wizzup | well I have newer u-boot I think | 22:17 |
| Wizzup | but others might need it | 22:17 |
| Wizzup | we don't have a way to auto update u-boot for folks | 22:18 |
| Wizzup | since there are so many possible setups | 22:18 |
| uvos | yeah | 22:18 |
| uvos | some solution is still needed | 22:18 |
| Wizzup | yes, but like with d4, I want to see the bits on my screen here first ;) | 22:18 |
| parazyd | Wizzup: Yeah, just append the dtb in debian/rules | 22:20 |
| Wizzup | freemangordon: with that commit reverted the corruption seems gone | 22:24 |
| freemangordon | ъеах | 22:24 |
| freemangordon | yeah | 22:24 |
| freemangordon | as expected (see the commit message :) ) | 22:25 |
| Wizzup | shall I revert it or drop it? | 22:25 |
| Wizzup | yeah | 22:25 |
| freemangordon | better drop | 22:25 |
| Wizzup | ok | 22:25 |
| Wizzup | rebuilding ddx | 22:26 |
| uvos | Package libllvm8 is not available, but is referred to by another package. <-- where is this supposed to be? | 22:51 |
| parazyd | We'll sort it out | 22:51 |
| parazyd | You can get it from beowulf-backports for the time being | 22:51 |
| uvos | backports ok | 22:51 |
| parazyd | (That's devuan upstream repos, not ours) | 22:52 |
| uvos | yeah i know thanks :) | 22:52 |
| parazyd | :) | 22:52 |
| uvos | y | 22:55 |
| uvos | upps wrong window | 22:55 |
| uvos | [ 43.960] (EE) OMAP(0): ERROR: failed to reconfig crtc 1 | 23:34 |
| uvos | hmm | 23:34 |
| uvos | what am i missing? kmscube works fine so dose weston | 23:34 |
| Wizzup | where di you get this? | 23:34 |
| uvos | xorg.log | 23:34 |
| uvos | it also [ 29.770] (II) OMAP(0): Successfully initialized the "omap_pvr" sub-module | 23:34 |
| uvos | so pvr ddk is working fine for video-omap too | 23:34 |
| Wizzup | it is harmful? | 23:35 |
| uvos | yes xorg dosent start | 23:35 |
| uvos | errors are fatal :P | 23:35 |
| uvos | by convetion anyways | 23:35 |
| Wizzup | I think I see it too but it doesn't seem to be fatal? | 23:35 |
| uvos | hmm | 23:36 |
| uvos | so i just rebooted | 23:38 |
| uvos | and now it did start | 23:38 |
| uvos | but its extreamly slow | 23:38 |
| uvos | like llvmpipe slow, but nothing changed in xorgs log | 23:38 |
| Wizzup | yes, this is mesa.sh | 23:39 |
| Wizzup | # cat /etc/profile.d/mesa.sh | 23:39 |
| Wizzup | #!/bin/sh | 23:39 |
| Wizzup | export MESA_LOADER_DRIVER_OVERRIDE=pvr | 23:39 |
| Wizzup | It might need to do +x I am not sure | 23:39 |
| Wizzup | this will be in sgx-ddk-um-tools momentarily | 23:39 |
| Wizzup | also the init script | 23:39 |
| uvos | huh | 23:39 |
| uvos | why is that needed | 23:39 |
| uvos | kmscube dose fine | 23:39 |
| Wizzup | what do you mean? | 23:39 |
| uvos | it uses sgx renderer | 23:39 |
| Wizzup | what is 'it' ? | 23:40 |
| Wizzup | the env var? | 23:40 |
| uvos | yes the env var, why are we forcing pvr | 23:40 |
| uvos | with kmscube mesa chooses pvr by its own will | 23:40 |
| Wizzup | because mesa trying to load the wrong driver name, I think it's been discussed ~5 times here in the last month or so | 23:40 |
| Wizzup | with this picked at the 'best' solution next to make a symlink | 23:40 |
| uvos | yeah except thats not what i see | 23:41 |
| Wizzup | in X? | 23:41 |
| uvos | kmscube -D /dev/dri/card1 | 23:41 |
| Wizzup | I don't know what path that takes | 23:41 |
| uvos | version: "OpenGL ES 2.0 build 1.17@4948957" | 23:41 |
| uvos | renderer: "PowerVR SGX 540" | 23:42 |
| uvos | Wizzup: should be same as x | 23:42 |
| Wizzup | try to set the env and see if X becomes accelerated? | 23:42 |
| uvos | yeah | 23:42 |
| uvos | sec | 23:42 |
| Wizzup | I needed it, before I did that it was llvmpipe too | 23:42 |
| Wizzup | parazyd: we'll need sgx-ddk-um-tools in some meta too | 23:42 |
| uvos | Wizzup: hmm yeah with the envvar its acellerated | 23:45 |
| uvos | but hmm | 23:45 |
| uvos | wierd why would mesa suddenly choose differently for the same dri device | 23:45 |
| uvos | 768 frames in 5.0 seconds = 153.477 FPS | 23:47 |
| uvos | 5x improvement | 23:47 |
| uvos | not bad | 23:47 |
| freemangordon | what is this? | 23:48 |
| uvos | gears | 23:48 |
| freemangordon | ah | 23:48 |
| uvos | firefox-esr dosent start | 23:48 |
| uvos | (segfaults instead) | 23:48 |
| * bencoh headscratches - 153fps?! | 23:49 | |
| Wizzup | surf does the same I think | 23:49 |
| bencoh | is it without vsync this time? | 23:49 |
| uvos | bencoh: its allways been without vsync | 23:49 |
| bencoh | ah, alright | 23:49 |
| freemangordon | yeah, no vsync on d4 | 23:49 |
| bencoh | I thought we somehow reached the conclusion that it was somehow synced but that it could manage a bit more than 60 | 23:50 |
| bencoh | but ... fair enough | 23:50 |
| uvos | that was on android | 23:50 |
| bencoh | ah | 23:50 |
| Wizzup | freemangordon: you also need sgx-ddk-um-tools now | 23:52 |
| Wizzup | it contains init script and mesa.sh | 23:52 |
| freemangordon | ok | 23:52 |
| uvos | there is no debug symbols for ff-esr in devuan | 23:53 |
| uvos | whats up with that | 23:53 |
| uvos | :( | 23:53 |
| Wizzup | dbg vs dbgsym? | 23:53 |
| freemangordon | debug symbols are in separate repo | 23:53 |
| uvos | ah yes | 23:54 |
| Wizzup | right | 23:54 |
| uvos | im an idiot | 23:54 |
| freemangordon | deb http://debug.mirrors.debian.org/debian-debug buster-debug main contrib non-free | 23:54 |
| uvos | thanks, yeah i just flashed a new sd card and ,... | 23:54 |
| freemangordon | np | 23:55 |
| Wizzup | so with -tools I think it 'just works' as in no manual changes required | 23:59 |
| Wizzup | now to sort out the dep | 23:59 |
| Wizzup | s | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!