I'll fix the uImage bodhi. I've been running off the internal sata using the factory load since yesterday and have run the transfer test script multiple times without a problem so far.by sodface - Debian
//note, I posted twice in a row, see my last post post on the previous page for the fail I had testing M300 #5. I need to fix the uImage for the bodhi kernel I'm using, apparently I made it with the 0x10008000 value I noted above: ## Booting image at 00100000 ... Image Name: Linux-5.5.1-kirkwood-tld-1 Created: 2020-04-10 15:36:18 UTC Image Type: ARM Linux Kernelby sodface - Debian
Well, I moved my test spinning hard disk over to M300 #5 and ran the test script before going out to walk the dog. Just got back and reviewed the results. Looked ok. Started the script again, below is what happened, md5sums at the beginning are the end of the first script run: m300-kernel-5.4.28-build-dir.tar.gz 100% 592MB 11.1MB/s 00:53 617c938c2053d6ebfac73397b729896bby sodface - Debian
Thanks for all the inputs gents. I'll do some more testing and update the thread. @LME, I have five M300's also. One I'm using as my main firewall/router running shorewall on Alpine connected to a managed switch (router on a stick type setup with the single M300 interface). It's been working great with the stock 16GB SSD: router:~$ uptime 15:42:46 up 21 days, 19:05,by sodface - Debian
>> Assuming that the stock filesystem and kernel was reliable MIke, I do have the old load still and did some brief testing with it and did _not_ see the issue though scp speed was much worse, I was only seeing ~6MB/s with the stock kernel/rootfs vice ~12MB/s with the newer loads, I don't think this is relevant to the issue, just a comment. I didn't do extensive testing withby sodface - Debian
Left the hard drive connected up to USB over night. I've run the test again a couple of times today, no issues. I have yet to duplicate the problem when booted from USB.by sodface - Debian
bodhi Wrote: > Do you do sync each time, and how long you wait > in between copying? No sync. The md5sum takes ~23 seconds to complete for each file, though if I run it again immediately on the same file it takes ~5 seconds so somewhere between 5-23 seconds minimum between transfers though really it's a lot more than that because the script md5sums all the files each time and Iby sodface - Debian
bodhi Wrote: > Try Ext3 with the same md5 test, see what'll > happen. I switched over to ext3 for the rootfs and wrote a little test script to scp the same file from the server 10 times, each time changing the local file name to get a new local copy and also running md5sum against all files before starting the next transfer. That completed successfully last night. I left the Mby sodface - Debian
I decided to test with bodhi's 5.5.1 kernel and 5.2.9 rootfs. I wanted to configure it closely to what I've been doing with Alpine, using a small ext2 partition for /boot and a second partition for /. I'm not changing any of the default M300 uboot environment variables so I've been setting the root=/dev/sda2 in the .dts file and then changing the kernel config to "Eby sodface - Debian
I'll see if I can make this clear and concise. I'm still having an issue with data integrity running on Alpine. Basically: - Boot off internal sata, login and immediately scp a ~500MB tar.gz file. - md5sum the file, it's correct. - scp the same file again to a new local name so I have two local copies of the same file. - md5sum the second file, it's correct. - cmp tby sodface - Debian
Hello again, I wanted to follow up on this thread with a couple of things related to the M300 (still with Alpine Linux). First is, I started having some issues with network transfers of larger files being corrupt (md5sums not matching, crc errors when extracting). The test file I was using was around 1.6GB, but I also had corruption in a 200MB file. I tried a bunch of different things but I&by sodface - Debian
bodhi Wrote: ------------------------------------------------------- > That's not the behavior in Debian as I known of. > But I guess it's all academic. Glad you got it > figured out what to do in Alpine! I'm still trying to figure out what the default behavior in Debian is supposed to be, based on Mike's output it looked like it default to 0 (ignore). If thatby sodface - Debian
bodhi Wrote: ------------------------------------------------------- > However, I think Debian is more correct in that > such applications bugs should be fixed. Unless you > can make it apparent so that users can submit a > bug report, it is like sweeping dirts under the > rug :) Besides, there is no guarantee that the > kernel could fix it up successfully. Some of the &by sodface - Debian
bodhi, while I agree with your statements, I think that the following is also true (even today) as far as the alignment setting goes: - The debian config you guys are currently using sets /proc/cpu/alignment to User faults: 0 (ignore) - Due to this setting, applications that perform unaligned memory operations may exhibit unexpected behavior because the kernel will not correct it and the progby sodface - Debian
Out of curiosity, I booted back up the factory Dell M300 image and checked the alignment setting: root@dellkaceM300:~# cat /proc/cpu/alignment User: 0 System: 0 Skipped: 0 Half: 0 Word: 0 DWord: 0 Multi: 0 User faults: 2 (fixup) So that's kind of interesting. It's not being set via the cmdline: root@dellkaceM300:~# cat /proc/cmdline console=ttyS0,115200 mtdby sodface - Debian
bodhi Wrote: ------------------------------------------------------- >We welcome any Linux > distro users. Thanks bodhi! > I did not recall Mike has "bus error" ? He didn't, but his /proc/cpu/alignment output from his m300 running LMS listed 3 alignment faults - no idea really how to interpret what caused them and as these are in the "system" categoby sodface - Debian
bodhi and dancyran, forgive me for being new to the forum and butting in here beating the same drum as in the m300 thread, but the timing is such that I feel I should post the link the mailing list thread again and the fact that the memory alignment issue gave me a a "Bus error" also. Coincidence? Perhaps? Perhaps not? https://lists.alpinelinux.org/~alpine/apk-tools/%3CCAGP1gyMH+Gby sodface - Debian
bodhi Wrote: ------------------------------------------------------- > sodface, > > This box architecture is Armv5te, i.e Kirkwood. > Same as other Kirkwood boxes that we have been > supporting since kernel 3.16.x. So I dont think > alignment is a problem at all. Hi bodhi, you guys have way more knowledge on this than I do so I don't doubt you are correct, at leby sodface - Debian
Thanks Mike, this memory alignment thing on the M300 may be something that could also affect you guys running Debian. See: https://wiki.debian.org/ArmEabiFixes#word_accesses_must_be_aligned_to_a_multiple_of_their_size and https://www.kernel.org/doc/html/latest/arm/mem_alignment.html It looks like by your output above that you are set to "0" which I believe is what my Alpine insby sodface - Debian
Thanks for the info mikeh49. I bought an Apple USB-C dongle/DAC and a two pack of USB-C to USB-A adapters from Amazon based on this review and some other positive comments about the Apple dongle: https://www.audiosciencereview.com/forum/index.php?threads/review-apple-vs-google-usb-c-headphone-adapters.5541/ It's the first Apple product I've ever purchased and I did it reluctantlyby sodface - Debian
bodhi Wrote: ------------------------------------------------------- > Nice use of the internal SATA port! Thanks. Next job is to get the bottom plate drilled out so I can mount the drive brackets and get it closed back up. Hopefully it doesn't get too hot in there! > That's a good alternative if you stay with this > system configuration and not booting other. Youby sodface - Debian
Hi all, I saw mikeh49's post over on the slimdevices forum which had a link back here and just wanted to say thanks for the info and work you've done on the m300. I bought 1 off ebay for $13.95 shipped (new in the box) and once I got it I felt like it was such a good deal I bought 3 more! I've been spending a lot of time with Alpine Linux and have been working on getting all thby sodface - Debian