Jump to content
XPEnology Community

IG-88

Developer
  • Posts

    4,628
  • Joined

  • Last visited

  • Days Won

    210

Everything posted by IG-88

  1. just to se it boot to the grub menu should work on any system (i assume 1.04b loader for 918+ with your hardware) most systems will indicate a "missing" ram with a beep code (if a small buzzer is present) when you now have monitor you should at least be able to see a boot screen of the systemboard or enter the bios (if the ram is working)
  2. a tv might not have all the graphics modes a s a monitor, so it might not show some screens like text mode in the grub.cfg selection if the system is set up the right way (usb boot passible) and the usb is prepared the right way it would be possible to set up the nas without a monitor (as its done on a original unit, they too have not monitor attached when setting it up), the usb should boot with the 1st option grub and the system should be found in network with the assitant
  3. and i guess as long as synology does not publish source of the kernel used for 6.2.3 it will stay like this (and even with source, as i can't code i would not be able to backport a newer i915 driver to synology's 4.4.59 kernel) the new update 2 for dsm 6.2.3 comes with a complete new set of compiled kernel modules so its to expect that the i915 patched driver is overwritten and /dev/dri will be missing after the update, the version of the new u2 driver is still the same (20171222), so it would need a newly patched version of the i915.ko (the "old" patched version might work when replacing the u2 version but the clean solution is to patch the new u2 version too)
  4. 16 ist der default in 1.04b sandisk cruzer fit oder ähnliche sind die einfachste lösung m.2 oder paltte gehen nicht, das ding simuliert das usb modul der orginal box und wird auch bei updates von dsm mit dem neuen kernel beschrieben, selbst wenn man es irgendwie hinbekommen würde müsste man auch sicherstellen das der mount prozess während des updates das richtige boot volume mountet zum updaten - so tief willst du sicher nicht in die sachen einsteigen, ein kleiner usb stick löst dein problem für ein paar euro wenn die platten die jetzt in deinem system sind ein dsm haben müsste das system ganz normal booten und die alten system partitionen ignorieren, evtl. wird auch eine warnung angezeigt, wenn es auf den daten partitionen der beiden platten kein nutzbares raid mit daten gibt die du übernehmen willst wäre es am einfchsten du löscht alle partitionen von den zwei platten, as vermeidet das irgendwelche schutz mechanismen aktiv werden die versuch dich davor zu ewahren eine evtl. wichtige platte mit daten drauf zu löschen
  5. ahh, good somehow i though it was the other way around, so it still seems like there is no case where it got worse when using the i915 driver that comes with 6.2.3 what loader and dsm type, what mainbaord and any additional hardware like other nic's?
  6. if the new synology i915 driver driver does not work for you then you only way would be downgrading to 6.2.2 and continue to use jun's i915 driver atm there is not even kernel source to see what synology did to backport there new i915 driver, its different from what jun did and jun's driver dous not work with 6.2.3
  7. could you try synologys videostation for transcoding with this patch? https://xpenology.com/forum/topic/24864-transcoding-without-a-valid-serial-number/
  8. vague reference, are you talking about 6.2.2 (jun's i915 driver) and 6.2.3 (synology's own newer i915 driver - initially made for the newer gemini lake x20 units) the "mod" i915 driver only contains a small patch to replace the hardware id of 8086:3E92 (9th gen CPU) with the id of newer cpu'd id using the "same" UHD630 iGPU, nothing related to your older iGPU is changed, in theory a fresh install with loader 1.04b and my newer 6.2.3 extra/extra2 (0.11 and newer) should result in a working /dev/dri on 6.2.2 there where two i915 versions the one from synology not supporting anything above apollo lake and jun's i915 made from newer source with support up to 9th gen cpu's with 6.2.3 synology replaced its older driver with a newer source similar to jun's version and might had a few more optimizations and tweaks to compensate some quirks in it so its working for all cases instead of the situation on 6.2.2 where the two i915 drivers had different sweet spots (like apollo lake better to be used with synology's driver) jun might have had a harder job (and limitations) with his i915 driver as he could not change the kernel, everything needed to be in the driver to fit synology's 4.4.59 kernel with 6.2.3 synology synology changed some things in its 4.4.59 kernel for a better fit of the i915 code they where using and broke jun's i915 driver on the way the result with 6.2.3 is still pleasing a all hardware now woks as before but with the stock driver from synology
  9. afair synology dropped wifi support ~3 years ago, the normal solution (beside connecting it to a existing network that has a wifi router) would be to use a access point as wifi client or bridge to connect to a wifi network there are also other technology's like PLC/PowerLAN you could use to connect to a existing network with wifi support wifi integration in linux was never really good and might be the reason synology dropped it, to much effort for a few percent of users, its mainly about usb wifi as most synology consumer devices dont have pcie slots with usb 3.x the speed might not be the problem but most usb dongles don't come with antenna connectors and relying on a tiny integrated antenna usually results in terrible wifi performance
  10. you were supposed to use the latest extra/extra2 with the 6.2.3. kernel and thats 0.13.3 and not 0.4,my link above was to that version
  11. use the extra/extra2 from the 1st post? if you come from a former version containing jun's extra/extra2 an his i915 driver this driver will not work with the kernel from 6.2.3 and the driver will also prevent the new one from synology to be loaded also in the 1st post is a list of supported pid's from synology's driver, 0412 is in that list there should be no i915.ko in /usr/lib/modules/update/
  12. Atom CE5315, "Berryville" (32 nm), ~2013? GPU PowerVR SGX545 - no support for using hardware transcodeing with juns laoder network: RTL8211E -> its just a phy chip so the network driver might be part the soc and intel i'd suggest trying a fresh install on a test disk first, if that work you can use the usb to install to your disks and complety replace the system, its not worth trying to "migrate" to keep same settings just install with the option to keep your "data" (raid with your data), system should be replaced make sure you test with the test install like copying data over network to see if performance is ok, if that does not work its much less work to just keep it the old way and replace the unit later with new hardware an a new dsm install
  13. that helps, i know whats wrong i updated the link in the 1st thread to a new "mod2" version, please download and try again
  14. no thats just the normal realtek 8168 driver and thats still part of the loader you could try to extract rd.gz and zImage from the dsm 6.2.3 *.pat file (918+) using 7zip, copy that to the loader (2nd partition) replacing the files there and then use the new extra/extra2 from here, also copy to 2nd parttion replacing the files there https://xpenology.com/forum/topic/28321-driver-extension-jun-103b104b-for-dsm623-for-918-3615xs-3617xs/
  15. netif_num is the number of network cards so it should be 4 in your case set netif_num=1 -> set netif_num=4 but i dont think thats the problem as jun checks later in the grub.cfg and sets it ... if common_add_option mac1; then set netif_num=1; fi if common_add_option mac2; then set netif_num=2; fi if common_add_option mac3; then set netif_num=3; fi if common_add_option mac4; then set netif_num=4; fi common_add_option netif_num ... what you get with "lspci -k"? there should be 2x igb for the i350T2 1 x igb for i211at (onboard) 1 x e1000e for i219v (onboard) jun did not set it to a higher value in the patch (inside extra.lzma), its the same as with the 12 drives in 3615/17 its not in the patch so it keeps the way as in the original but you can change manually and this change will be kept as long as the synoinfo.conf is not replaced, like when installing a bigger update (200-300MB *.pat file), in this case all things in jun's patch will kick in so the 16 drives will work instantly but your manuall changes (like 4 nic's in 918+ or >12 drives in 3515/17) will be gone and need to added again (i has this on my system when i was going from 6.2.3 25423 -> 25426 (both where "full" dsm versions) the following is directly copyed from my synoinfo.conf and i have 3 working nic ports maxlanport="4" eth2_mtu="1500" eth2_wol_options="d" eth3_mtu="1500" eth3_wol_options="d" looks the same as your changes so the problem must be something else thats not new at all it was like that from the beginning, my warning about that (you linked above) is two years old
  16. my delete link was still valid, so they must have a interesting way to determine the 100 days for not downloading anyway its up again with a new link
  17. means wrong version number of kernel, you would need to change the makefile synology delivered https://xpenology.com/forum/topic/7187-how-to-build-and-inject-missing-drivers-in-jun-loader-102a/?do=findComment&comment=122631
  18. i tried with docker and jellyfin but hardware acceleration with nvidia nvenc had no effect on playback in native resolution and when encoding was active it did not work at all maybe its because it needs some special docker tools we don't have with dsm's docker https://jellyfin.org/docs/general/administration/installing.html#hardware-transcoding-with-nvidia-ubuntu a native software might work but i will not pay for plex or emby just to try something i dont even use maybe i will make a test version and a little howto, i'm not even sure if there is a need for this if its not possible to use without coughing up extra money to use hardware transcoding when most replay devices can cope with 4k streams
  19. steht in der faq, die anzeige in dsm ist fix und zeigt nur das an was im original war/wäre es wird genurtzt was der kernel hergibt, die wesetliche begrenzung ist da 8 kerne bei 3615 und 16 kerne bei 3617 wenn es 2nd gen ist dann geht 918+ nicht, board wäre nur wichtig wegen legacy, ahci mode für sata und was für ein onboard nic es hat
  20. not everyone build that "small" i'm using a i3 9100, the 8 cores of the kernel is starting to be a limit you can have a look at my newer 6.2.3 extra's for 918+, i removed the i915 driver from the package and modded the patch to delete the old one when present (like when updating from 6.2.x to 6.2.3) + +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/agpgart.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/cfbcopyarea.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/cfbfillrect.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/cfbimgblt.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/drm.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/drm_kms_helper.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/fb.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/fb_sys_fops.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/i915.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/intel-agp.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/intel-gtt.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/syscopyarea.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/sysfillrect.ko +/tmpRoot/usr/bin/rm -f /tmpRoot/usr/lib/modules/update/sysimgblt.ko +
  21. glaube ich erher nicht, leider hast du vergessen zu schreiben was für ein board und welchen i5 du verwendest wenn es haswell i5 4xxx oder neuer ist kannst du auch 1.04b mit 918+ nehmen, 1.03b und 3615/17 booten nicht bei reinem uefi, da braucht es csm/legacy mode und man muss dann auch das legacy usb device wählen (da man evtl. erst sieht wenn man einmal mit akrivem csm/legacy gebootet hat) https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/
  22. i guess your i915 driver for hardware transcoding is not working that way (/dev/dri) also jun's patches in extra and extra2 are different ahh, so it was you, not the people who adapted the patch in the first place why? it does not hurt in any way to have more "slots" in the web gui, it will only limit the usability for other people i'd tend to up the 16 to 24 as this is a tested number of disks and will be enough in 99.9% of the cases
  23. die erst mal wird nur versucht das raid (und das lvm2 volume sofern vorhanden) , bei problemen bricht das ab, beim mounten wird -ro wie in "read only" verwendet so das duc nicht mal versehtlich was von dem gemounteten volume löschen kannst sowas in der art gibts auch hier im forum https://xpenology.com/forum/topic/7004-tutorial-how-to-access-dsms-data-system-partitions/
×
×
  • Create New...