Welcome to The Linux Corner!

Discussion in 'The Linux Corner' started by Deleted member 368765, Feb 27, 2017.

Thread Status:
Not open for further replies.
  1. I own a Razer Blade Stealth and Core 2017 edition, both were shipped to me in the last two weeks. The issues I am encountering are these:

    1. The laptop will randomly hang if you press the Capslock key. This is a huge error in the current Linux support. To my understanding this is affecting all Kabylake hardware. Intel/Kernel developers need to get this fixed.

    2. The Razer Core does not currently work. This could be due to several factors, and I will be borrowing a modern GPU shortly to try to test if the GPU is the problem (770GTX at the moment) lspci is showing the gpu as connected but I can't bring the drivers up properly.

    3. The community Keyboard drivers are good but could use some cleaning up, documentation would probably help them improve the keyboard drivers to support all the lighting options, the Core and the Keyboard both have correct lighting.

    4. BIOS Updater for Linux. There is an AMI BIOS updater utility, but it does not build/run cleanly on most distributions. A Razer provided BIOS updater would be helpful. Creating a UEFI Boot disk is a complete mess and the laptop appears to be unable to boot freedos from a USB. Currently only Windows can run the BIOS updates (YMMV)
  2. Excalibur4212

    Excalibur4212 New Member

    Oh my gosh. This is great! This forum will probably help with my issues with wine!
  3. The inability to upgrade BIOS natively from linux is deal killer. The reality is that it does not matter if Razer's linux drivers only work with Razer's hardware, as long as they work.

    The next release of the PCs below should see me taking the plunge, because the ports should be upgraded to non-legacy. (if you don't put the spec of the I/O slot on your website, it's legacy) and linux working from the install disks.

    Why the desire for 2 SSD's? Because we are linux users. Many of us routinely set up dual/multiple boots etc. However, there is risk to the drive used for O/S's. It is a LOT safer to keep data on a separate physical drive and just blast a bare metal image back onto the SSD used for O/S's from a fast USB stick if things go South, and be up and running in a few minutes, with not a bit lost.

    My decision is down to three laptops:
    • Razer Blade Stealth
      It doesn't have a second SSD drive. The graphics are good, it's light, at 70 WHr it has a good sized battery and it looks cool, but it's not a zero bezel.

    • Dell XPS 15.
      A little research shows that it has two battery sizes. The 90WHr battery takes the space of a full 2.5" drive. If I want two drives, I have to go with the 60 WHr battery one SATA drive. With a zero bezel, it's a 14" in size, and has a 4k display, and is a little heavier. The ports are decidedly legacy. But it seems to have less compatibility problems with linux than Razer. It can have up to 32GB of RAM.

    • Sager NP3130.
      It has two drives optional, both PCIe. The QHD display at 3200 x 1800 is not as high res of graphics as the two above, but better than HD. The battery is small at 36WHr, but it is as light as the Razer. Its only stated spec for the display port is 1.3. It can have up to 32GB of RAM. It's decidedly not a zero bezel. I'm not sure how compatible with linux it is, but by the next release it will be determined.
    Can anyone from Razer give timing on status of linux bios updates from linux and drivers working on install ?
    Last edited: Apr 4, 2017
  4. I own a Razer Blade Stealth and Core 2017 edition, both were shipped to me in the last two weeks. The issues I am encountering are these:

    1. T̶h̶e̶ ̶l̶a̶p̶t̶o̶p̶ ̶w̶i̶l̶l̶ ̶r̶a̶n̶d̶o̶m̶l̶y̶ ̶h̶a̶n̶g̶ ̶i̶f̶ ̶y̶o̶u̶ ̶p̶r̶e̶s̶s̶ ̶t̶h̶e̶ ̶C̶a̶p̶s̶l̶o̶c̶k̶ ̶k̶e̶y̶.̶ ̶T̶h̶i̶s̶ ̶i̶s̶ ̶a̶ ̶h̶u̶g̶e̶ ̶e̶r̶r̶o̶r̶ ̶i̶n̶ ̶t̶h̶e̶ ̶c̶u̶r̶r̶e̶n̶t̶ ̶L̶i̶n̶u̶x̶ ̶s̶u̶p̶p̶o̶r̶t̶.̶ ̶T̶o̶ ̶m̶y̶ ̶u̶n̶d̶e̶r̶s̶t̶a̶n̶d̶i̶n̶g̶ ̶t̶h̶i̶s̶ ̶i̶s̶ ̶a̶f̶f̶e̶c̶t̶i̶n̶g̶ ̶a̶l̶l̶ ̶K̶a̶b̶y̶l̶a̶k̶e̶ ̶h̶a̶r̶d̶w̶a̶r̶e̶.̶ ̶I̶n̶t̶e̶l̶/̶K̶e̶r̶n̶e̶l̶ ̶d̶e̶v̶e̶l̶o̶p̶e̶r̶s̶ ̶n̶e̶e̶d̶ ̶t̶o̶ ̶g̶e̶t̶ ̶t̶h̶i̶s̶ ̶f̶i̶x̶e̶d̶.̶
    This has been fixed in Kernel 4.11. As distributions pick up the new kernel the issue will go away.

    2. T̶h̶e̶ ̶R̶a̶z̶e̶r̶ ̶C̶o̶r̶e̶ ̶d̶o̶e̶s̶ ̶n̶o̶t̶ ̶c̶u̶r̶r̶e̶n̶t̶l̶y̶ ̶w̶o̶r̶k̶.̶ ̶T̶h̶i̶s̶ ̶c̶o̶u̶l̶d̶ ̶b̶e̶ ̶d̶u̶e̶ ̶t̶o̶ ̶s̶e̶v̶e̶r̶a̶l̶ ̶f̶a̶c̶t̶o̶r̶s̶,̶ ̶a̶n̶d̶ ̶I̶ ̶w̶i̶l̶l̶ ̶b̶e̶ ̶b̶o̶r̶r̶o̶w̶i̶n̶g̶ ̶a̶ ̶m̶o̶d̶e̶r̶n̶ ̶G̶P̶U̶ ̶s̶h̶o̶r̶t̶l̶y̶ ̶t̶o̶ ̶t̶r̶y̶ ̶t̶o̶ ̶t̶e̶s̶t̶ ̶i̶f̶ ̶t̶h̶e̶ ̶G̶P̶U̶ ̶i̶s̶ ̶t̶h̶e̶ ̶p̶r̶o̶b̶l̶e̶m̶ ̶(̶7̶7̶0̶G̶T̶X̶ ̶a̶t̶ ̶t̶h̶e̶ ̶m̶o̶m̶e̶n̶t̶)̶ ̶l̶s̶p̶c̶i̶ ̶i̶s̶ ̶s̶h̶o̶w̶i̶n̶g̶ ̶t̶h̶e̶ ̶g̶p̶u̶ ̶a̶s̶ ̶c̶o̶n̶n̶e̶c̶t̶e̶d̶ ̶b̶u̶t̶ ̶I̶ ̶c̶a̶n̶'̶t̶ ̶b̶r̶i̶n̶g̶ ̶t̶h̶e̶ ̶d̶r̶i̶v̶e̶r̶s̶ ̶u̶p̶ ̶p̶r̶o̶p̶e̶r̶l̶y̶.̶
    This works, the NVIDIA driver has to be bound to the PCI ID of the card in the Xorg.conf file.

    3. The community Keyboard drivers are good but could use some cleaning up, documentation would probably help them improve the keyboard drivers to support all the lighting options, the Core and the Keyboard both have correct lighting.

    4. BIOS Updater for Linux. There is an AMI BIOS updater utility, but it does not build/run cleanly on most distributions. A Razer provided BIOS updater would be helpful. Creating a UEFI Boot disk is a complete mess and the laptop appears to be unable to boot freedos from a USB. Currently only Windows can run the BIOS updates (YMMV)
  5. lkcl_

    lkcl_ New Member

    then don't use nvidia for video playback! i learned this lesson quite quickly with the Aorus X3PlusV6: http://lkcl.net/reports/aorus_x3_plus_v6.html

    like you, initially i thought "oh wow! NVidia Graphics! That'll Be Cool And Fast To Do Video Including Using OpenGL, RIght?? Optimus Is Cool So Let's Use It and Expect Everything To Be Fast, Right??"

    utterlly, utterly and completely wrong. like... so wrong in every single way that, after now having sorted out using the intel vaapi acceleration and DRI2 direct-to-screen rendering and having around 30% CPU usage (if that) at the lowest (800mhz) clock rate on 720p videos, i'm kinda surprised how utterly wrong i got it.

    here's a simple mpvrc file which will give you both VAAPI for decode and VAAPI for video display:

    then (if you use debian) apt-get install i965-va-driver and mpv should already be capable of picking that up because of its dependency libva1

    that's it - that's all you need. it's a similar story for VLC, although i've not found the video output mode which allows VAAPI to be used, at the very least for goodness sake use it for decoding:

    apparently i'm using xcb_xv in VLC:

    which still gives well under 30% CPU usage.

    likewise vlc should pick up the i965-va-driver by way of its dependency libva1. if you want to test things use vainfo (apt-get install vainfo) - generally you should be able to find precise resources online about this. DON'T use vdpau.

    you should also be setting some options in xorg.conf although you will need to experiment with this, currently i am using these

    Option "AccelMethod" "sna"
    Option "TearFree" "true"
    Option "DRI" "3"

    basically don't for god's sake expect OpenGL to do any good, here: not on the Intel Graphics and certainly not on the NVidia Graphics. all that will happen is it will eat CPU cycles.

    use the right hardware acceleration for the right job, and you should do okay.
    Last edited: Apr 8, 2017
    EverBot likes this.
  6. I'm totally in agreement for Razer to create its own Razer Linux distro. @Pongstr I would totally appreciate you weighing in on this post https://insider.razerzone.com/index.php?threads/razer-linux-distro.21758/
    Last edited: Apr 9, 2017
  7. Her0Zer0

    Her0Zer0 Member

    Thanks for this info! I haven't had a chance to fully digest it yet, but clearly you have quite a bit more depth on this subject that I do. However, I will say two things:

    1. I have NEVER gotten Optimus to work in any capacity on my 2016 Razer Blade. In fact, every time I've tried to mess around with Bumblebee, I ended up bricking the whole thing and had to simply reinstall. So that's out and I won't even bother with it (though do wish it did work). Thankfully, PRIMUS has always worked, though obviously the having to logoff/logon just to switch is a big hassle.

    2. I never expected the dGPU to be good for video playback. However, there's a NOTICEABLE difference using the dGPU for media heavy web browsing - with the obvious exception for video playback, it can handle far more open tabs of media heavy pages in Firefox. Of course, #1 makes this an annoying compromise (though it's improved a bit).

    To follow up on my original post, the latest nVidia drivers have brought significant improvements. Specifically, the iGPU has a BIG improvement offered - the ability to power down the dGPU. Engaging this option has increased battery life for me by 66% with just a button press - I'm easily getting 5+ hrs on battery where I would be lucky to get 3 before. Of course it was all of two days before the whole setup got screwed on reboot. Thankfully, after about 20 hrs of diagnosis, I finally figured out the problem has to do with bbswitch correctly loading/unloading the nvidia kernel modules when switching modes (ie. basically rmmod all the nvidia kernel modules from non-graphical boot if it is locking up when using iGPU).

    So two points to the broader thread I want to make:

    HARDWARE - I think this topic should be limited to Linux support of components as opposed to what people are saying they would prefer as a configuration. I totally understand the varying desires people have, but it muddies the discussion to talk about desired hardware changes. That being said, hardware support is a VERY important area - especially regarding Synapse and peripherals. I recently got a Razer Mamba and am absolutely LIVID I can't use the side scroll because the only OSS drivers out there are for an older model that dont have side scroll. However, those drivers (Polychromatic) DO allow some medicore level of control of LED lighting - basic effects and the Mamba actually works wirelessly with them. So it's better than nothing but far from sufficient - I like the Mamba but $150 and not having sidescrolling is infuriating, especially given that my main use is coding. As far as how Razer should go moving forward, I personally don't want ANY hardware sacrifices made for "Linux" models - I want nVidia, I want the quad core cpu, I want the full resolution (I use an LG 31in 4k monitor and the laptop drives it very well from thunderbolt). Just my 2 cents.

    SOFTWARE - lots of good debate in this thread about distros, repos, window managers, etc. I personally use Ubuntu Unity because it's probably the best support flavor out there and easy to find fixes for. I definitely think focus should be on Debian flavors for the same reason, though critical is planning a move from X11 and unfortunately, Weyland is supposed to be far superior to Mir (rumors that Ubuntu is going to abandon the latter, from what I've read). systemd has been trial and error, but I've thus far mostly made my peace with it. Finally, security is a critical component - passwords, ssh configs, fail2ban, ufw, etc. Linux is substantially more powerful as a server versus Windows, which makes it a constant target from hackers. I've spent weeks customizing my machines for security, but I also had a horrible breach only 6 months ago - and it happened from stupidly installing compromised deb files from the web. Obviously, in comparison to Windows, Linux security is miles above - but its also easy to forget that it's clearly not fully secure and user error can create an even worse scenario than compromising a Windows machine (ie. botting every PC on your LAN/WLAN).

    I really hope Razer pushes forward on this front - specifically, getting all the hardware supported first and foremost. The latest Ubuntu distro (16.10) works pretty well out of the box, but chroma, touchpad, peripherals, GPU, and power management need to be tuned - some more than others (ahem MAMBA - before I go insane). Given that most of the hardware already has FOSS drivers out there, I think packaging them up and regularly updating them for the desired distros would be an ideal place to start this endeavor. That process will undoubtedly lead to many more ideas about OS level things that would be nice. But perhaps, even before that, Razer needs to settle on distro(s) to support (and perhaps which hardware). These first two steps can be started immediately and I strongly believe its one of those cases where this SHOULD be started before the full vision is apparent. Because for quite a few of us, it already works. If we can do it, I have no doubt Razer can do it better and cleaner.
  8. lkcl_

    lkcl_ New Member

    you didn't "brick" the machine. "bricking" only occurs with embedded devices that have NAND Flash as the primary boot device, which becomes corrupted so badly that the device requires opening up with a screwdriver in order to consider recovering it.

    intel machines have an SPI-based NAND boot loader which contains the BIOS. if that got corrupted by Bumblebeed that would be *extremely* serious.

    what you must be referring to is the following:

    * machine booted up fine (BIOS ok)
    * grub bootloader booted up fine
    * OS started fine
    * OS loaded bbswitch module
    * OS tried to start X11 on top of bbswitch and SPANGGGG blank screen.

    you didn't even get to the X11 login because xorg locked up the machine, solid, right? (i know what's going on, basically, because i had the same thing happen and had to investigate it).

    now, i have an advantage of *always*, always *MANUALLY* starting X11 with the command "startx". i log in at the TEXT console, then once that is successful i run "startx" and run the session that *i* want. because i suspend the machine rather than power it off it's only if there's a problem that i have to go through this process again (last "login" was beginning of march, at the time of writing).

    so, from the text console, i did some investigation. i used "strace -o log.txt startx" to find out what the hell is going on, and the *moment* that xorg tried to read the /sys/class/devices/pcie/....../nvidia/config file, SPANGGGG

    this "wtf" moment led me to conclude that the bbswitch kernel module is clearly written by idiots, because *without* the bbswitch module loaded, X11 started up just fine. i also tried telling xorg to *not* do scanning of the PCIe bus... but it *completely* ignored my wishes even though the xorg config option clearly stated "don't scan the PCIe bus".

    anyway, next was to try loading the nvidia kernel module *before* starting xorg... and amazingly that worked perfectly.


    your mileage may vary: if you want to get bumblebeed up and running you will need to do some investigation, including doing things like switching OFF automatic starting of xorg so that you can log in at the console, experiment, and go from there: or, you can work out what kernel parameters / boot options will get you into console-only (but not single-user mode) - i don't know what distro you run and if it's ubuntu DON'T ask me for help... :)

    ... but basically you should investigate adding "nvidia" to /etc/modules - that really should be enough. that's a debian-specific thing, btw, for sysvinit: you'll have to track down the equivalent for your distro.

    if that doesn't work, make use of /etc/rc.local (or equivalent). i wrote a little startuphack script as follows:


    /etc/init.d/bumblebeed stop
    rmmod snd-hda-intel
    rmmod iwlmvm
    rmmod iwlwifi
    modprobe nvidia
    modprobe iwlwifi
    modprobe snd-hda-intel

    basically as you can see, i ran into problems with sound and wifi as well as the nvidia kernel module. *after* you start x11 *then* you can start bumblebeed (from a terminal), and optirun should work perfectly fine.

    basically optirun is considered experimental, and you can see why :)

    try the above.

    i'm running 375.26 and to be absolutely honest i've given up on nvidia. it's a constant battle that adversely affects my workflow. 375.20 was unstable and WOULDN'T WORK with a 4.8 or above kernel so i had to run with 4.7 for a while. they fixed that... only to introduce instabilities which cause the module to segfault... and take out the hardware with it! the only recovery mechanism is a full reboot.

    proprietary drivers being developed by proprietary idiots basically never goes well - not for something as complex as the world's largest software project (the linux kernel).

    if i *absolutely must* have access to the GPU i'll try again (and not look forward to doing so), but to be absolutely bluntly frank i'll take stability at the expense of a little less battery life and a little less speed any day.

    i explained on the other thread why this is not recommended: it's too much for a hardware company to do. the absolute simplest viable approach is to do a "Golden Image" that's factory-copied (by non-experts) onto the SSD and it's shipped. done. simple. and also the *user's* responsibility if they want anything more fancy.

    lastly, everbot: i won't be unkind because you've clearly learned the lesson now, but *what were you thinking* downloading arbitrary .debs??? this is *not* windows, with "DLL hell": that problem was solved *two decades* ago with package managers and distros! aptitude exists for a reason, not least is the GPG-signing on the packages *AND* the proper, stable management and installation of dependencies! if you install a .deb it f***s up your system by placing you in the position of then having to track down the dependencies... which now you can't install using aptitude or apt-get because your system's now in a broken state!

    for god's sake don't download arbitrary .deb packages - get the damn source code and build it yourself. if the source code *isn't available* then don't install it *at all*! anyone who doesn't provide the source code is looking to entrap you - period. the method of entrapment will vary, but it's entrapment all the same.
  9. Speaking of newest standards. WIFI is not always available. Even if Razer doesn't put a cell phone PCIe card that can accept SIMs into their laptops, it would be nice if Razer DID put the antenna for cell phone PCIe cards in case we want to add such a card later.
    This is NOT a promo, just an example (and it appears to have a linux option) http://www.elecomes.com/goods.php?id=180
  10. therave39

    therave39 New Member

    Hey, I'm currently looking into picking up a 2017 Razer Blade for my new work laptop, and was planning to install Ubuntu on it.

    I've been looking around and have seen the following issues exist:
    * External Monitors don't work
    * Graphics switching is a headache
    * Chroma doesn't work
    * Webcam doesn't work

    I'm not sure how out of date the info I've been looking at is, so have any of these been resolved?
  11. lkcl_

    lkcl_ New Member

    this situation can often be solved by installing the latest (experimental) mesa utils and kernel packages - obviously with the proviso that it's experimental. at the time of writing, mesa 17 is known to support dual display 2560x1080 external monitors, for example.

    the best distro for doing that (keeping always up-to-date) is going to be archlinux. their wiki is also fast becoming one of the best online resources for getting stuff running... *even on other distros*.

    i don't know what this is (and you didn't provide a link) so can't advise

    on the Aorus X3 Plus V6 the webcam is a UVC video device. it works straight out-of-the-box. it's also a really *good* UVC video device, providing 720p by default.

    so... a lesson there for Razer... to not f*** about with proprietary hardware that requires proprietary drivers. for example i never, ever got the camera working on my macbook pro.... because f****** apple used the totally unethical company broadcomm's products.

    do you have the output from "lsusb" and "lspci" available so that the camera can be identified?

  12. therave39

    therave39 New Member

    Hey, sorry I didn't provide more details! Chroma would be the application controlling the RGB lighting on the laptop.

    All of the things I mentioned weren't working were documented on the Ubuntu Wiki page for the Razer Blade, I just wasn't sure what's changed since that was last updated

    Right now everything is still research for me, I'm in the market for a new development laptop and greatly prefer Linux distros over MacOS
  13. Armydrifter

    Armydrifter Active Member

    As a Linux user for my Army job a lot. I would love this and be a beta tester of this on a new Razer Blade pro! I need something to practice on for my job as well as keep up to date on IT. I am certified on a few different vendors and getting more very soon. So something like this would be an amazing future addition!
  14. jwatte

    jwatte New Member

    One of my engineers installed Debian Testing for work on a GTX 1060 4k Razer Blade 14" (had to get the previous version 'cause the new version isn't available yet,) and she was quite happy that the multi-touch panel worked in Chrome on Linux :)
    She uses Bumblebee for the Optimus integration.
    Just reporting a happy Linux customer :)
    Hactar88 likes this.
  15. The issue with legacy ports is the number of PCIe lanes in the chipset.
    The HM175 intel chipset used in the upcomming Razer Pro only has 16 pcie lanes.

    However, the intel Z270 chipset has 24 pcie lanes. That's 8 pcie lanes more. That is to say enough for two more pcie x4. That could give the additional 3.1 gen2 port and an additional m2 pcie x4 ssd. Say, in the next Razer Blade?
  16. Mr.ShoNuff

    Mr.ShoNuff New Member

    I read something earlier regarding the need for good battery life.... well, I remember buying the first Asus ROG laptop.. the G1 or something (back in 2007). I was getting 4 hours on battery... little less playing WoW. The key to good linux battery life is don't rely on the vanilla kernel and when you just install Ubuntu and don't compile your own kernel you just have to deal with what comes with the distro with no performance benefits other than what is provided out of the box.

    Can't really beat the NVIDA gpu and their proprietary drivers on linux.. no matter what anyone says. That is changing though... when the AMD Ryzen and Vega are made mobile for latops... then maybe AMD will be viable --> definitely waiting for the AMD switch... their benchmarks as of late have been very impressive.
    Last edited by a moderator: Apr 20, 2017
  17. jwatte

    jwatte New Member

    You really need at least one regular USB port, because people carry USB sticks around in pockets and on desks and whatnot. I don't mind a ton of type C; that'd be OK; but any computer that doesn't have any type A at all is going to be a pain to actually use "in real life" for the next few years.
  18. illathon

    illathon New Member

    I am really happy to hear this. Thanks a lot. As a developer that needs CUDA, using Razer products are a good choice.
  19. Agreed on the form factor, one USB standard port, but it can be 3.0.

    Also need the SDXC card reader as they can store a bootable image and fit into a wallet. If the design of the SDXC port is like the old Lenovo laptops, it pushes in with a detent/click and is almost flush with the laptop, can the card can stay there. You can use the bios to decide whether or not you want to boot from it.
  20. This is where the Core could really shine, providing both NVIDIA and AMD graphics output capability.
Thread Status:
Not open for further replies.
Sign In with Razer ID >

Don't have a Razer ID yet?
Get Razer ID >