IRC Logs for #crux Tuesday, 2019-12-31

*** Kruppt has quit IRC00:56
*** mrwoggle has quit IRC01:28
dlcusaI just updated the xfce_4.14 repo for exo, tumbler, and xfce4-terminal--details available at http://dlcusa.net/CRUX/xfce_4.14 -- let me know of any problems, please.01:42
*** zimmer_Dl has joined #crux01:55
*** emmett1 has joined #crux02:09
*** emmett1_ has joined #crux02:43
*** emmett1 has quit IRC02:46
cruxbot[xorg.git/3.5]: libglvnd: updated to version 1.3.002:58
*** utilz has quit IRC03:08
*** utilz has joined #crux03:10
*** tilman_ has joined #crux04:46
*** beli3ver has joined #crux06:03
beli3verGood morning06:03
*** beli3ver has quit IRC06:06
*** emmett1_ has quit IRC08:21
*** emmett1_ has joined #crux08:22
cruxbot[compat-32.git/3.5]: libglvnd-32: 1.2.0 -> 1.3.008:58
*** abenz__ has joined #crux09:09
*** abenz_ has quit IRC09:12
*** emmett1_ has quit IRC10:08
*** joacim has joined #crux11:11
*** obarun has quit IRC12:32
*** pedja has joined #crux14:00
*** zimmer_Dl has quit IRC14:32
*** zimmer_Dl has joined #crux14:35
*** beli3ver has joined #crux15:13
beli3verHi can somebody help me with a grub efi installation15:14
beli3verI have created two partition. 1 is / with ext4. The second is /boot/efi fat32.15:18
beli3verI compiled the kernel and did15:19
beli3vercp arch/x86/boot/bzImage /boot/vmlinuz15:19
beli3vercp System.map /boot15:19
beli3verThen15:19
beli3vergrub-install /boot/efi15:19
beli3vergrub-mkconfig -o /boot/grub/grub.cfg15:20
beli3verNo15:20
beli3verErros15:20
beli3verBut when I restart I get the grub recovery screen15:21
jaegerIf you get the recovery screen that means grub is installed and loaded but it can't find its config file, probably15:22
*** weednix has quit IRC15:22
*** Kruppt has joined #crux15:23
*** weednix has joined #crux15:24
beli3verSo the grub mkconfig is the problem?15:28
jaegerIn the recovery shell can you find your root partition and the kernel via the builtin ls and tab completion?15:29
jaegerI suspect it's just the kernel and System.map name in this case15:30
jaegerIf I remember right, grub-mkconfig wants to see vmlinuz-* or kernel-*15:30
jaegerI typically name them vmlinuz-<version> and System.map-<version> and those work well15:31
jaegeryeah, search for 'vmlinuz' in /etc/grub.d/10_linux if you want to see the various strings15:31
beli3verOK thanks I will give a try15:32
joacimcan also modify the grub scripts so they can find vmlinuz and vmlinuz.old15:33
beli3verLol grub is searching for vmlinuz-*15:34
joacimor modify the script make install runs so that it populates /boot with filenames grub expects to see15:34
beli3verIt is working thanks guys15:38
jaegernp15:40
*** utilz has quit IRC15:41
beli3verOK next error.15:49
beli3verThis is my fstab for the root15:49
beli3ver/dev/nvme0n1p3 / ext4 defaults,noatime,nodiratime,discard,errors=remount-ro 0 115:49
beli3verWhen I start get kernel panic could not found the root device. I checked it 5 times. The device is correct15:49
beli3verLLol15:54
jaegerIf you're sure the device is correct then make sure the kernel has everything properly selected... in this case NVME support, SCSI disk support, ext4 filesystem support, etc.15:54
beli3verMissed kernel15:54
beli3ver:l15:54
beli3verSorry15:55
jaegerAs a side note, I think the discard option isn't needed for NVME drives15:55
beli3verJo15:56
beli3verYea15:56
beli3verYes15:56
ckrzen@bel16:00
ckrzen * beli3ver: I've read a lot about the 'discard' option causing problems. Better to just enable a scheduled TRIM(fstrim) job.16:01
beli3verYeah gave me a second boot error without it my surface pro 6 is running crux16:01
ckrzen * beli3ver: I've read a lot about the 'discard' option causing problems. Better to just enable a scheduled TRIM(fstrim) job. Once-a-week or so.16:01
jaegerdiscard is still fine (in my experience) for non-NVME SSDs, just not recommended for NVME16:04
*** heroux has quit IRC16:08
ckrzenok...just re-checking my assumptions, it seems that the 'discard' option only frees-on-erase commands(implying that the option needs to be in fstab asap after formatting?) and that fstrim can still be periodically called to check that there aren't any unreclaimed blocks that the 'discard' option missed(at least on recent SSD firmwares).16:25
*** heroux has joined #crux16:30
jaegerI believe that's the case but I'm not an expert there16:45
*** c0rnelius has joined #crux16:56
cruxbot[contrib.git/3.5]: python3-babel: 2.7.0 -> 2.8.017:03
beli3verIs there a iwlwifi package17:07
jaegerThere's linux-firmware if you're looking for firmware for an iwlwifi card17:08
beli3verI need support for a Marvell 88W889717:10
jaegeryou should be able to see exactly which files it wants via 'dmesg | grep firmware'17:11
ckrzenbeli3ver: jaeger here's the relevant info on NVMe and SATA TRIM that I was remembering:17:14
ckrzenNVM Express17:14
ckrzenThe NVM Express command set has a generic Dataset Management command, for hinting the host's intent to the storage device on a set of block ranges. One of its operations, deallocate performs trim. It also has a Write Zeroes command that provides a deallocate hint and allows the disk to trim and return zeroes.17:14
ckrzenACTION sent a long message:  < https://matrix.org/_matrix/media/r0/download/matrix.org/GZxkwANKslcCLVvbEYwVHsiG >17:15
beli3verI get this error for the firmware17:16
beli3verhttps://s19.directupload.net/images/191231/w5suhyvw.jpg17:16
ckrzenACTION sent a long message:  < https://matrix.org/_matrix/media/r0/download/matrix.org/INmiZFRRtFYzvxOGjNselkTQ >17:16
beli3verI have no idea what to do17:16
ckrzenhttps://en.wikipedia.org/wiki/Trim_(computing)#Disadvantages17:17
jaegerstart with 'prt-get fsearch pcie8897_uapsta.bin'17:17
jaegerThe i915 ones are for the intel GPU17:17
beli3verOK will give a try17:19
*** zimmer_Dl has quit IRC17:51
*** zimmer_Dl has joined #crux17:56
*** beli3ver has quit IRC18:03
*** beli3ver has joined #crux18:03
beli3verLol forget to install linux-firmware18:14
beli3ver:)18:14
*** beli3ver has quit IRC18:18
*** beli3ver has joined #crux18:19
ckrzenAny real-world observations for using "-march=native" vs "...=x86_64" in pkgmk.conf if I'm just building on  a couple of recent vintage laptops(i5-4xxx/5xxx)?18:46
jaegerMy only comment is that -march=native will occasionally cause weird issues or failed compiles... not very common but does happen... so if you run into an issue building something the first thing I'd suggest is turn off native and try again18:48
ckrzenkool...thanks.18:48
ckrzen"-O2" vs O3, etc?18:50
jaegerhaven't played with those much18:51
ckrzenk. thankx18:51
*** zimmer_Dl has quit IRC18:57
*** c0rnelius has quit IRC18:58
*** BurnZeZ has joined #crux19:01
*** BurnZeZ has quit IRC19:06
*** zimmer_Dl has joined #crux19:34
*** BurnZeZ has joined #crux19:35
*** zimmer_Dl has quit IRC19:36
*** Anselmo has joined #crux19:49
stenurA good year 2020 i wish everybody.  (Already now, ciao ciao ciao!!)19:55
*** Anselmo has quit IRC21:20
*** obarun has joined #crux21:42
ckrzenAm I correct in understanding that "prt-get" will work without "ports -u" or is it necessary to always or, at least, periodically perform a "ports -u"?21:50
jaegerports -u is still necessary21:50
ckrzenk21:51
uplimeeven for a sysup, a ports -u is recommended first21:56
*** beli3ver has quit IRC22:01
*** beli3ver has joined #crux22:01
abenzckrzen: fwiw, I have march=native since I started using crux, no issues whatsoever22:22
abenzfrom some benchmarks online (phoronix), targeting native could yield some noticable performance gains22:23
abenznoticeable*22:24
ckrzenabenz: Great! Thanks.22:26
ckrzensysup'ing my install for the first time with:22:26
pedjaI am a bit sceptical about phoronix benchmark 'methodology'22:27
ckrzenCFLAGS="-Ofast -march=native -pipe". fingers-crossed!22:28
pedja-OGiveMeAllYouGotScotty22:28
ckrzenAye!22:29
pedjaI bet that line from ST is a basis for some fanfic :)22:29
pedjawill I look it up? hell no22:30
ckrzenOh...yah!22:30
*** Kruppt has quit IRC22:55
pedjahappy new year, humans :)23:06
pedjaand other intelligent life forms23:08

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!