#
✕ nick:arnd
#asahi · 2021-02-15 13:50
arndmodwizcode: it certainly does, you have to trim that manually, but that is the same as with any other client
#asahi · 2021-02-15 13:48
arndmodwizcode: the address I send from redirects to gmail, so I just use that for replying. It's not great, but I got stuck with it after a previous job required using it, and I have never started using mutt
#asahi · 2021-02-15 13:47
arndI used to have a script that takes adds everyone who added a Reviewed-by/Acked-by/Signed-off-by/etc tag to the original patch referred to by a 'Fixes: abcd12345678 ("commit shortlog")' line, but now this is done automatically
#asahi · 2021-02-15 13:45
arndgit-send-email will leave out the duplicates from cc
#asahi · 2021-02-15 13:45
arndand the with 'to' I only pick the names that are the people that I expect to eventually apply the patch, and with Cc I pick whatever seems most appropriate
#asahi · 2021-02-15 13:43
arndI tend to use 'git send-email --to-cmd="./scripts/get_maintainer.pl -i --no-git --no-git-fallback --m --no-r --no-l" --cc-cmd="./scripts/get_maintainer.pl -i --git --r --l"'
#asahi · 2021-02-15 13:42
arndmarcan: there is no good automated way. Best apply some common sense to limit the Cc's to the ones that are listed as maintainers and the people with the most commits. It's definitely easier once you know who everyone is ;-)
#asahi · 2021-02-15 13:10
arndkrzk of course should review whether the changes are ok before that
#asahi · 2021-02-15 13:09
arndmarcan: gregkh will have to provide an Ack for merging the changes through the soc tree instead of the serial tree
#asahi · 2021-02-15 12:30
arndmarcan: you will probably get a semi-automated message from gregkh, telling you that he doesn't review serial driver patches during the merge window
#asahi · 2021-02-15 12:28
arndwe rarely have one that requires changes to core code like this one does
#asahi · 2021-02-15 12:27
arndthen again, we sometimes see new SoC support from maintainers that already own multiple other platforms, and those might go in after v1. Same for anything that is just a minor revision of an existing SoC family
#asahi · 2021-02-15 12:26
arndrarely, most new maintainers are less thorough in their initial bringup
#asahi · 2021-02-15 12:22
arndthe main thing now is to get Acks from the arm64, irqchip, dt and serial maintainers so I can merge it through the soc tree
#asahi · 2021-02-15 12:22
arndthere are probably a few very minor comments that can just be addressed in without a resend
#asahi · 2021-02-15 10:13
arndbut it was clearly thrashing while I tried to work inside of the 12GB machine, without anything interesting going on outside
#asahi · 2021-02-15 10:13
arndthere was also a lot of building all of compiling enough homebrew to run a patched qemu, and whatever the other user did
#asahi · 2021-02-15 09:17
arndtarzeau: this was with MacOS running with one user logged in graphically and another user running a text-only Linux guest through ssh most of the time
#asahi · 2021-02-15 08:01
arndI had tried running a 12GB Linux guest on the 16GB mac mini, which ended up completely unusable. Giving it 8GB was fine mostly. I have about 1TBW after 19 hours of total machine usage (both according to smartctl)
#asahi · 2021-02-14 19:05
arndnever_released: this my threadripper workstation that has pretty much been building kernels for the past 3.5 years
#asahi · 2021-02-14 19:04
arndsmartctl says "Data Units Written: 295,530,779 [151 TB]"
#asahi · 2021-02-14 19:01
arndIIRC it's vendor specific, I should check the datasheet
#asahi · 2021-02-14 18:59
arndMy four year old Samsung 960 Pro using MLC is at 23% with 295529554 "data units written"
#asahi · 2021-02-14 18:58
arndI wonder if they used MLC flash then, most drives today use either TLC or QLC, which ends up with a much shorter life
#asahi · 2021-02-14 18:57
arnddavidrysk[m]: it's also really RAM-limited
#asahi · 2021-02-14 18:55
arndor "nvme smart-log /dev/nvme0n1"
#asahi · 2021-02-14 18:55
arnddavidrysk[m], Glanzmann: looking at https://www.heise.de/preisvergleich/?cat=hdssd&xf=18284_2000~4832_3&asuch=&bpmin=&bpmax=&v=e&hloc=at&hloc=de&hloc=pl&hloc=uk&hloc=eu&plz=&dist=&fcols=4930&bl1_id=300&sort=-filter4930#productlist, the best 2TB drives are advertised as with a lifetime 4.4PBW, the cheapest ones only 320TBW
#asahi · 2021-02-14 17:05
arndmarcan: before you send out v2, can you add a patch to 'defconfig' to enable the platform and any driver you need?
#asahi · 2021-02-14 17:03
arndmarcan: right, I actually had found the same Kconfig bugs independently but never got around to sending out the fixes (I had two patches instead of one, but it makes sense to combine them as you did)
#asahi · 2021-02-14 15:39
arndright. For the opposite case (new kernel, old dtb), there is usually an easier way to avoid breaking things, but it may require doing the update over multiple merge windows, to first change the driver and then dte dts files
#asahi · 2021-02-14 15:37
arndeven if only one way is broken, it can still be a pain for users to upgrade when they want to use a newer dtb to use new features that a later kernel enables, but after updating the dtb they can no longer go back to a known working kernel
#asahi · 2021-02-14 15:36
arndthere are lots of other things that could go wrong though
#asahi · 2021-02-14 15:35
arndit's possible that is no longer the case, it's been a while since I looked at iommus
#asahi · 2021-02-14 15:35
arndmarcan: I think if the dtb lists an iommu, the kernel will not try to use devices until a driver for the iommu has been loaded, but older kernels would then fail because they don't have a driver for it
#asahi · 2021-02-14 15:26
arndthe main problem later on is when you have a .dts file in the kernel and intentionally break compatibility with older versions. Which is something that we try to avoid as soon as there are users anyway, but it might happen with the iommu or something unexpected
#asahi · 2021-02-14 15:25
arndmarcan: yes, we specifically allow this for the initial merge to avoid having platform maintainers go through multiple release cycles to get everything merged in the right order
#asahi · 2021-02-14 15:14
arndafter the initial merge, the second method is normally used for everything
#asahi · 2021-02-14 15:13
arndthere are generally two options for merging a new platform: either everything goes into the soc tree, with an ack from the other subsystem maintainers, or everything goes into separate trees, and only the dts+Kconfig+MAINTAINERS file updates go through soc.git
#asahi · 2021-02-14 15:12
arndmarcan: just to clarify the timeline: I expect linux-5.11 to be released later today, which means it's too late for 5.12 now. I won't queue anything up until v5.12-rc1 is out, but I'm happy to take anything that has been reviewed at that point into soc.git for v5.13, which means it can be part of linux-next two to three weeks from now
#asahi · 2021-02-14 12:51
arndmarcan, robher: if we assume the new property could show up in any node, it makes sense to document it there, OTOH it's likely that no other platform will need it any time soon
#asahi · 2021-02-14 12:49
arndhttps://github.com/devicetree-org/devicetree-specification/blob/master/source/chapter2-devicetree-basics.rst is where dma-coherent is documented
#asahi · 2021-02-13 19:14
arndnever_released, kettenis: I see the internal pci has three interrupts from the host bridge, while the thunderbolt has just one interrupt line, but neither of them has an interrupt-map property in adt, so I wonder how one should find the correct interrupt lines
#asahi · 2021-02-13 14:41
arndthis is highly unlikely to work, but if someone wants to find the bugs and fix them, feel free to take that as a starting point
#asahi · 2021-02-13 14:39
arndI wrote up a prototype for an nvme-platform driver, based on marcan's latest test branch https://git.kernel.org/pub/scm/linux/kernel/git/arnd/playground.git/log/?h=nvme-platform
#asahi · 2021-02-12 12:13
arndmarcan: if you have a custom firmware for that device, do you think it also act as a software programmable i2c slave on the DDC? That might be enough to either fake up an HID keyboard over i2c, or something that can work as a text console, given a little code on each end of it.
#asahi · 2021-02-12 11:07
arndearly ixp4xx have the almost the same exact problem as H6
#asahi · 2021-02-12 11:07
arndwe also have something in arch/arm/mach-ixp4xx/ that does it in a less portable way, and I'm hoping to fold that in there as well
#asahi · 2021-02-12 11:06
arndjhol: johill is working on an abstraction for readl/writel that he wants to use on arch/um to fake a PCI bus, the same should work on Allwinner H6
#asahi · 2021-02-12 10:49
arndI'm still trying to figure out if there is some way to pretend to be a keyboard remotely
#asahi · 2021-02-12 10:47
arndmaz: in the modern world we live in, a 4K HDMI-to-USB screengrabber is actually cheaper than a 1.2V usb-serial cable, but I get your point ;-)

plus ›