Re: [gentoo-user] System crash, more than once. Out of memory problem.
On Mon, Nov 26, 2018 at 6:16 PM Alan Grimes wrote: > > Programmers these days don't seem to have any concept that memory is > finite and that it's usage should be optimized. =| > I have it on good authority that unused memory is wasted memory. It is a good thing that I never run more than one process a time so that my browser is correct in assuming that the OS wouldn't possibly be able to find a better use for the RAM than that tab I haven't clicked on in 20 minutes that is running some javascript with more lines of code than the sorts of operating systems people actually used back when they didn't run more than one process at a time... -- Rich
Re: [SOLVED] Re: [gentoo-user] Plasma sound devices
On Sunday, 25 November 2018 16:24:21 GMT Peter Humphrey wrote: > On Saturday, 24 November 2018 10:06:37 GMT Peter Humphrey wrote: > > On Friday, 23 November 2018 19:28:19 GMT Dale wrote: > > > Peter Humphrey wrote: > > > > Hello list, > > > > > > > > Over the last few days I've trimmed my plasma setup - to slightly > > > > below > > > > the > > > > minimum, by which I mean that sound notification has stopped working. > > > > I > > > > can > > > > hear streamed radio, but the plasma control panel doesn't show any > > > > devices > > > > so I don't hear notifications. > > > > > > > > I've scanned the list of packages I removed but I can't see a likely > > > > suspect, and google just returns results about having no sound at all, > > > > which is not my case. > > > > > > > > Can anyone tell me which plasma module I need to install to hear > > > > notifications again? > > > > > > > > Again, the hardware and low-level drivers are fine: it's just plasma > > > > that's > > > > missing a component. > > > > > > Could you be missing Kmix? Since it controls volume/mute settings, I'd > > > think it would be needed, maybe anyway. > > > > Good idea, but no cigar. > > > > Mick suggested pulseaudio. I don't think it can be that because there's no > > sign of it in my backups; neither in world nor under /etc/portage. I'll > > try > > it though; I just need to decide whether to set the USE flag globally or > > just on certain packages - or just emerge pulseaudio. The global setting > > causes 10 rebuilds, including ffmpeg and firefox, as well as 9 new > > packages, so don't hold your breath. > > That didn't help either. In fact I had to install phonon-gstreamer. > > I hope that helps somebody one day. Interesting ... I don't have phonon-gstreamer, but media-libs/phonon with USE="pulseaudio vlc", which brings in media-libs/phonon-vlc here. Either way, one or the other phonon backend seems to be necessary to be able to produce sound. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] System crash, more than once. Out of memory problem.
Dale wrote: > Howdy, > > Things I've done based on what I've noticed while sitting at the > system. I noticed with the newer Firefox versions, the multi-process > ones, that it can hog up a TON of memory. Sometimes I wonder if it is > trying to download the entire internet or something into memory. One > Firefox profile can be eating up several GBs of memory even with few > active tabs. To help with that, I changed the process count from 4 to > 1. That seemed to help with that a LOT. It hasn't gotten to hungry > since. Basically back to the way it was before the multi-process thingy. Amen, bro. Programmers these days don't seem to have any concept that memory is finite and that it's usage should be optimized. =| -- Please report bounces from this address to a...@numentics.com Powers are not rights.
[gentoo-user] RaptorCS Blackbird - Owner controlled, open source firmware system on the POWER ppc64/ppc64le arch - a less expensive mATX TALOS 2
This is a much less expensive mATX variant of the TALOS 2 from the same people. It runs both little and big endian so both ppc64 and ppc64le. https://www.phoronix.com/scan.php?page=news_item=Blackbird-POWER9-Pre-Orders https://raptorcs.com/content/BK1B01/intro.html The only binary blob is the NIC firmware[1], otherwise it is fully open source and has no hardware code signing enforcement so it is entirely yours unlike modern x86 stuff which can't ever be free[2] and has the impossible to disable ME/PSP doing god knows what. OpenPOWER9 CPU's are Made in USA and the board is Made in the USA from US and foreign components so it is much more trustworthy. In terms of speed POWER9 is superior to the offerings from intel/amd or equivilant without x86's spectre/meltdown protections enabled which intel usually dishonestly performs in their benchmarks. [1]It was the best alternative to using an intel nic as there is a large amount of documentation available, people are working on freeing it and the first one to do so gets a free TALOS 2 workstation. [2]New x86 hardware has none of the documentation published that is required to write firmware and it has a variety of black boxes like ME/PSP, boot guard etc designed to prevent you from owning and controlling your hardware.
[gentoo-user] System crash, more than once. Out of memory problem.
Howdy, I've ran into this more than once here lately. It's starting to get on my nerves. As some know, I reboot when power fails and generally not before. If I'm not sitting at my system, it's downloading stuff, usually videos or something. Plus I watch TV from it as well. At this point, something has a problem. I don't know for sure what but have suspects. Things I've done based on what I've noticed while sitting at the system. I noticed with the newer Firefox versions, the multi-process ones, that it can hog up a TON of memory. Sometimes I wonder if it is trying to download the entire internet or something into memory. One Firefox profile can be eating up several GBs of memory even with few active tabs. To help with that, I changed the process count from 4 to 1. That seemed to help with that a LOT. It hasn't gotten to hungry since. Basically back to the way it was before the multi-process thingy. The other day, after a upgrade, I noticed dolphin started triggering something. Dolphin starts to consume memory and then the X itself, the actual command X, starts to eat up memory. The only way I can get X itself to release it is to logout, restart the GUI from a console and then login again. If I see it quick enough, I can close the Dolphin window and restart it. Got lucky a time or two. I'm not sure but I use the little preview thingy to check downloaded videos with. I suspect that may be the trigger. I think F11 activates it. I'm attaching the relevant part of my messages file. Does anyone see anything in there that gives a clue? I see where X was a hog and that is the process the kernel killed, which of course, killed pretty much everything else. Thanks much. Dale :-) :-) P. S. I have more memory on the way but this is still a issue that needs fixing. If it eats all I have now, it will just eat more of it later if this is some sort of bug. Nov 26 13:10:01 fireball CROND[25787]: (root) CMD (/usr/lib64/sa/sa1 1 1) Nov 26 13:11:22 fireball kernel: http.so invoked oom-killer: gfp_mask=0x6200ca(GFP_HIGHUSER_MOVABLE), nodemask=(null), order=0, oom_score_adj=0 Nov 26 13:11:22 fireball kernel: CPU: 2 PID: 25049 Comm: http.so Tainted: P O 4.18.12-gentoo #1 Nov 26 13:11:22 fireball kernel: Hardware name: Gigabyte Technology Co., Ltd. To be filled by O.E.M./970A-UD3P, BIOS F1 08/06/2013 Nov 26 13:11:22 fireball kernel: Call Trace: Nov 26 13:11:22 fireball kernel: dump_stack+0x46/0x59 Nov 26 13:11:22 fireball kernel: dump_header.isra.33+0x72/0x286 Nov 26 13:11:22 fireball kernel: ? preempt_latency_start+0x2e/0x41 Nov 26 13:11:22 fireball kernel: oom_kill_process+0x7e/0x31c Nov 26 13:11:22 fireball kernel: out_of_memory+0x3a2/0x3da Nov 26 13:11:22 fireball kernel: __alloc_pages_slowpath+0x4e5/0x634 Nov 26 13:11:26 fireball kernel: __alloc_pages_nodemask+0xfe/0x18b Nov 26 13:11:26 fireball kernel: ? find_get_entry+0x80/0x88 Nov 26 13:11:26 fireball kernel: filemap_fault+0x23b/0x329 Nov 26 13:11:26 fireball kernel: ? filemap_map_pages+0x67/0x228 Nov 26 13:11:26 fireball kernel: ext4_filemap_fault+0x28/0x3c Nov 26 13:11:26 fireball kernel: ? ext4_page_mkwrite+0x2fa/0x2fa Nov 26 13:11:26 fireball kernel: __do_fault+0x18/0x39 Nov 26 13:11:26 fireball kernel: __handle_mm_fault+0x33f/0x6b1 Nov 26 13:11:26 fireball kernel: handle_mm_fault+0xa1/0xa6 Nov 26 13:11:26 fireball kernel: __do_page_fault+0x20c/0x394 Nov 26 13:11:26 fireball kernel: ? page_fault+0x8/0x30 Nov 26 13:11:26 fireball kernel: page_fault+0x1e/0x30 Nov 26 13:11:26 fireball kernel: RIP: 0033:0x7f1341619fa0 Nov 26 13:11:26 fireball kernel: Code: b6 08 84 d2 0f 84 40 04 00 00 48 8b 7c 24 08 31 db 90 c1 e3 04 48 ff c7 01 d3 89 d8 25 00 00 00 f0 74 07 89 c2 c1 ea 18 31 d3 <0f> b6 17 f7 d0 21 c3 84 d2 75 dd 84 c9 74 34 48 8b 54 24 18 66 90 Nov 26 13:11:26 fireball kernel: RSP: 002b:7ffe02015ff0 EFLAGS: 00010246 Nov 26 13:11:26 fireball kernel: RAX: RBX: 004e RCX: Nov 26 13:11:26 fireball kernel: RDX: 004e RSI: 7f1342f93015 RDI: 7f13388c65d1 Nov 26 13:11:26 fireball kernel: RBP: 7f13388c65d0 R08: R09: Nov 26 13:11:26 fireball kernel: R10: R11: R12: 7ffe020161b0 Nov 26 13:11:26 fireball kernel: R13: 55a9a5af0f30 R14: R15: 0001 Nov 26 13:11:26 fireball kernel: Mem-Info: Nov 26 13:11:26 fireball kernel: active_anon:3452341 inactive_anon:565021 isolated_anon:0\x0a active_file:781 inactive_file:821 isolated_file:0\x0a unevictable:2846 dirty:0 writeback:0 unstable:0\x0a slab_reclaimable:13929 slab_unreclaimable:12120\x0a mapped:1420626 shmem:1914327 pagetables:19921 bounce:0\x0a free:20934 free_pcp:0 free_cma:0 Nov 26 13:11:26 fireball kernel: Node 0 active_anon:13809364kB inactive_anon:2260084kB active_file:3124kB inactive_file:3284kB unevictable:11384kB isolated(anon):0kB isolated(file):0kB mapped:5682504kB
