Krzysztof Kozlowski
Posts
210Following
33Followers
231#standwithukraine 🇵🇱 🇪🇺 🇺🇦 🇨🇭
IRC: krzk
Kernel work related account. Other accounts of mine: @krzk@mastodon.social
@conor Yes, this too. Several cases can be explained with non-obvious requests from the reviewer or just misunderstanding. Therefore I prepared few longer, detailed template answers with comments and instructions. The insult is when such long and explanatory reviewing comment is being ignored...
BTW, I find templates (Quicktext for Thunderbird, macros for mutt) or AI as the only way to stay sane. :)
BTW, I find templates (Quicktext for Thunderbird, macros for mutt) or AI as the only way to stay sane. :)
0
0
3
@kernellogger @gregkh But you know, adding proper sysfs interface takes few more lines of code… and one has to write documentation… and then sysfs is counted as ABI, so one cannot freely change it… and long command lines look so readable, just consider this Android-related stuff added by bootloader for one SoC:
androidboot.verifiedbootstate=orange androidboot.keymaster=1 androidboot.vbmeta.device=PARTUUID=815a78c5-ebb4-48ec-2912-90eeb3f71c48 androidboot.vbmeta.avb_version=1.2 androidboot.vbmeta.device_state=unlocked androidboot.vbmeta.hash_alg=sha256 androidboot.vbmeta.size=12800 androidboot.vbmeta.digest=815a78c5fb66af47291265ab90eeb3f71c48c9af6ad65044a08a5d81e77299a1 androidboot.veritymode=enforcing androidboot.veritymode.managed=yes androidboot.bootdevice=1184000.ufs androidboot.fstab_suffix=default androidboot.boot_devices=soc/1184000.ufs androidboot.serialno=19111140 androidboot.mode=ffbm-00 quite androidboot.baseband=msm msm_drm.dsi_display0=qcom,mdss_dsi_vtdr6130_fhd_plus_cmd:config0:timing0: androidboot.force_normal_boot=1 silent_boot.mode=nonsilent
Isn’t it beautiful?
2
2
3
Krzysztof Kozlowski
krzkGo Beavers! I mean, the true beavers, the greatest ecosystem engineers, which are vital in bringing diversity, richness, and healing degraded landscapes. “Beaver settlements triple in 15 years in Switzerland” https://www.swissinfo.ch/eng/beavers-increase-threefold-in-15-years-in-switzerland/48657934
If you want to learn more, why beavers are important, I recommend Andrew Millison’s: https://youtu.be/43bmtqKDhBE
0
1
3
Krzysztof Kozlowski
krzk
With v6.5-rc1 merge window, my remaining few fixes for Samsung Exynos SoC dtbs_check compliance were merged. This means that with v6.5-rc1 all ARM and ARM64 Samsung Exynos SoC Devicetree sources (DTS) pass in-kernel Devicetree bindings compliance tests. I can finally enable a builder testing mainline DTS for this:
https://krzk.eu/#/builders/92
https://krzk.eu/#/builders/91
https://github.com/krzk/tools/commit/743f694cfe4d18dd5f92728967aa762edd59c84e
P.S. The ARM64 Exynos DTS was actually compliant since v6.3-rc1, so that part I could have enabled earlier.
https://krzk.eu/#/builders/92
https://krzk.eu/#/builders/91
https://github.com/krzk/tools/commit/743f694cfe4d18dd5f92728967aa762edd59c84e
P.S. The ARM64 Exynos DTS was actually compliant since v6.3-rc1, so that part I could have enabled earlier.
0
5
14
Krzysztof Kozlowski
krzk
Edited 1 year ago
Is fixing the same issue over and over again, and expecting it to be fixed, also definition of insanity? Or just Groundhog Day? Or Whack-A-Mole game?
https://lore.kernel.org/linux-devicetree/20230711063011.16222-1-krzysztof.kozlowski@linaro.org/T/#u
https://lore.kernel.org/all/20230410175232.22317-1-krzysztof.kozlowski@linaro.org/
https://lore.kernel.org/linux-devicetree/20230711063011.16222-1-krzysztof.kozlowski@linaro.org/T/#u
https://lore.kernel.org/all/20230410175232.22317-1-krzysztof.kozlowski@linaro.org/
1
0
3
@cas @samueldr Patch based workflow is not necessarily easy, although there are tools making it easier (like b4 mentioned by @cas), but it is the one which is the easiest for us. Now, I have been sending patches with git send-email for some years and I have sent quite a few (if you read LWN then you know :) ), and I never had single problem with git send-email. Neither with git format-patch. I find them trivial. Now and long time ago.
I have been also working with people complaining a lot about Git, that it is too complicated and too difficult. I would say all of these - Git is too difficult, git send-email is too complicated - is the same complaining as "C is too difficult" or "using shell is too complicated": valid statements for a regular user, but not really for software developer in this domain. When learning something new, it is pretty often difficult and complicated. This is why it is called learning...
I have been also working with people complaining a lot about Git, that it is too complicated and too difficult. I would say all of these - Git is too difficult, git send-email is too complicated - is the same complaining as "C is too difficult" or "using shell is too complicated": valid statements for a regular user, but not really for software developer in this domain. When learning something new, it is pretty often difficult and complicated. This is why it is called learning...
0
0
2
Krzysztof Kozlowski
krzk
Hey @linuxfoundation
Do you plan to upload more photos of EOSS 2023? I see pictures from only the first two days of EOSS, not even complete two days... Other days are not uploaded at all. As you can see in the link below, count stays at 150 pictures is since last Thursday:
https://www.flickr.com/photos/linuxfoundation/albums/72177720308904909/page2
#eoss2023 #elce2023 #EmbeddedOSSummit
Do you plan to upload more photos of EOSS 2023? I see pictures from only the first two days of EOSS, not even complete two days... Other days are not uploaded at all. As you can see in the link below, count stays at 150 pictures is since last Thursday:
https://www.flickr.com/photos/linuxfoundation/albums/72177720308904909/page2
#eoss2023 #elce2023 #EmbeddedOSSummit
0
0
2
@shanmukhateja I think the post under link listed what was present in initial submission. The Adreno GPU patches were recently posted here: https://lore.kernel.org/all/20230628-topic-a7xx_drmmsm-v1-0-a7f4496e0c12@linaro.org/
0
0
1
Krzysztof Kozlowski
krzk
Edited 1 year ago
10% of commits in Linux kernel v6.4 came from Linaro engineers. It is a nice result for such a small group of kernel engineers.
https://www.linaro.org/blog/linaro-contributions-to-the-6-4-linux-kernel-release/
https://www.linaro.org/blog/linaro-contributions-to-the-6-4-linux-kernel-release/
0
1
8
Krzysztof Kozlowski
krzk
Edited 1 year ago
It was great to see at the EOSS 2023 commitments from some ARM64 SoC manufacturers to upstreaming their SoCs. It was visible in the conference talks but also I got such clear vibes during the talks behind the scenes.
What was a bit missing is a story of Qualcomm which recently is heavily visible upstream - both directly and through the Qualcomm Landing Team in Linaro. November last year Qualcomm released its newest, shiniest mobile SoC: the Snapdragon 8 Gen 2 (SM8550).
And guess what happened? The next day you could find all the Linux kernel support for this new SoC on the mailing list. The newest SoC was released with upstream support at the same time. There is also story from @superna9999: https://www.linaro.org/blog/upstream-linux-support-now-available-for-the-the-qualcomm-snapdragon-8-gen-2-mobile-platform/
#eoss2023 #elce2023 #EmbeddedOSSummit
What was a bit missing is a story of Qualcomm which recently is heavily visible upstream - both directly and through the Qualcomm Landing Team in Linaro. November last year Qualcomm released its newest, shiniest mobile SoC: the Snapdragon 8 Gen 2 (SM8550).
And guess what happened? The next day you could find all the Linux kernel support for this new SoC on the mailing list. The newest SoC was released with upstream support at the same time. There is also story from @superna9999: https://www.linaro.org/blog/upstream-linux-support-now-available-for-the-the-qualcomm-snapdragon-8-gen-2-mobile-platform/
#eoss2023 #elce2023 #EmbeddedOSSummit
1
15
39
@minipli @gregkh @kernellogger Timelines indeed sometimes do not align. Consider Ubuntu 18.04 (Bionic) which was using v4.15 kernel because of the timeline and Canonical was basically doing LTS on their own for four years. But timing cannot justify distro/vendor staying on chosen kernel for 4 years, instead of going to next LTS. It's distro's or vendor's deliberate choice - they don't want to update the kernel in released distro version even once per few years. And that choice is here being challenged.
1
0
1
@kernellogger Been there. Request to backport few KVM patches for new AMD processors to distro v4.15 kernel resulted in 500-big patch bomb... and I was only half way. Then we gave up and managed to convince customer that it's not a feasible solution.
1
0
6
Krzysztof Kozlowski
krzk
Edited 1 year ago
Room full of people - I certainly didn't expect such big audience. #eoss2023 #elce2023 #EmbeddedOSSummit
1
0
7
@kernellogger @mupuf the forges are not the answer. We already have tools which report such failures (LKP/kernel test robot), so either the tree was not on kernel.org (thus not covered by LKP by default) or reported warning was ignored.
What we maybe miss is a tool or part of workflow, which will check for how long patch was in next, before sending to Linus. And loudly complaining if it's less than two weeks.
What we maybe miss is a tool or part of workflow, which will check for how long patch was in next, before sending to Linus. And loudly complaining if it's less than two weeks.
1
0
0
Krzysztof Kozlowski
krzk
@monsieuricon xkcd nicely pointed out silliness of usual password approaches - difficult for us to remember, easy for machines to break: https://xkcd.com/936
0
0
0
Krzysztof Kozlowski
krzk
Updating my CV, collecting references, debt register extract, salary details and writing motivation letter. People do crazy stuff to get a nice apartment in Switzerland.
1
0
1
Krzysztof Kozlowski
krzk
I love it. And think how many other equations and laws of physics can be improved with adding AI. Ohm's law? Dude, please, I = (V + AI) / R. Newton's second law? Easy peasy, F = m * a * AI. Maxwell equations - don't get me even started...
2
0
3