linux on the hp envy 15
Computer folks, especially Linux hackers, are a little weird when it comes to vacation. A lot of them look forward to a large uninterrupted block of time so they can focus on their personal programming projects. The reason this is weird is because to the observer, the form of what the hacker is doing remains unchanged from normal: hunched over a keyboard, drinking a beverage of dubious quality from a can, muttering to self, occasionally pausing to scratch.
But inside the mind, it’s more fun than no-pants day at EPCOT center. Personal projects are fun precisely because they’re personal. You’re creating (or fixing) something just for yourself, and in the end, you get to reap the benefit of your own work. And if you can share the results with others, so much the better. It’s like you got to eat 10 funnel cakes by yourself and everyone else grows fat with you. Can’t beat that with a stick.
I’ve spent the past week or so playing with the HP Envy, which apparently is one of the jawesomest laptops on the market right now. Unfortunately, Linux wasn’t working that well on it and folks started complaining. Via some internal HP magic, I managed to get my hands on one to try and improve the situation (in my own time, as a volunteer, hence the lead-in to this post).
So here’s the sitch. I’ll update this page as I learn more.
First off, you need to be sure you have the latest BIOS. As of this writing, the latest revision is F.07.
I’ve been doing all my testing with a combination of Ubuntu (Karmic Koala) and upstream Linus kernels. If I get the time, I may play with some other distros, but for now, it’s all I’ve done.
what works
- Thermal controls — fans, CPU throttling, thermal zones etc. All that stuff is good to go; you won’t end up with a pool of dripping silicon and plastic burning into your crotch
- TurboBoost — the Intel Core i7 chip has a TurboBoost feature; I’ve confirmed that it works on the Envy
- just about everything else: wifi, sound, media card reader, etc.
fixes in-flight
- I submitted a patch series upstream to resolve an ACPI issue that some people were seeing in dmesg. The message was obtuse, but it was indicating that we couldn’t initialize the ambient light sensor. My patch re-enables it. Update:
(12/22) patch set applied to acpi-test.(12/25) patches in 2.6.33-rc2 - Takashi Iwai wrote a patch to enable the clickpad. Alternatively, you can workaround the lack of right-click by using the keyboard’s right-click button (next to the right control key) or use an external mouse.
- Update:
(1/20) Fixed how Linux handles C states, machine now idles much cooler in AC mode. patches will be in 2.6.33 (when released)Update: (2/9) backported to Ubuntu Karmic; more details below
workarounds
- Update: (2/14) Verified a workaround for screen corruption after suspend/resume. You need to create a minimal xorg.conf in /etc/X11/ that has the following contents:
Section "Screen"
Identifier "Default Screen"
DefaultDepth 24
EndSection
Section "Device"
Identifier "Default Device"
Driver "ati"
Option "EXANoDownloadFromScreen" "true"
EndSection
Copy and save the above into /etc/X11/xorg.conf, then issue sudo /etc/init.d/gdm restart and you’re good to go.
distro watch
Ubuntu Karmic
- I filed this bug to backport the C state fix into Karmic: ACPI: enable C2 and Turbo-mode on Nehalem notebooks on A/C. It has now been applied to the Karmic tree, so if you’re running Ubuntu, you’ll get the fix Real Soon Now ™. Lucid will automatically pick up this update too.
ugly warts
corrupt screen after resume. haven’t had time to work on this one yet. You can workaround this by issuing /etc/init.d/gdm restart after resume, but of course that will restart the X server and you’ll lose your work. Still, it may be faster than a full reboot.Update: (2/14) Discovered Launchpad bug screen corruption with Radeon 4770 and verified that the fix in comment #5 works on the Envy 15. The workaround makes the screen corruption after suspend/resume go away. Yay!
topics of extreme confusion
TurboBoost
TurboBoost requires support from the BIOS and hardly any involvement from the OS (whether it’s Windows or Linux). The Envy BIOS does indeed turn TurboBoost on. Yet, when you look in /proc/cpuinfo and look for “Hz”, you may only see 1.6GHz or 1.7GHz.
What is going on?
The issue is the way that the Linux kernel calculates the speed of the CPU. In summary, it does not display the speed 100% accurately all the time on CPUs with TurboBoost, and it probably never will. The reason I say it probably never will is because in order to calculate the speed of a TurboBoost CPU, the kernel would have to poll the hardware once every second. This is a pretty wasteful thing to do; your machine could be doing something useful instead of figuring out exactly how fast the CPU is running.
So you cannot simply look at the Hz output from the kernel in order to determine if TurboBoost is on. In fact, the way you do so is by looking at another place in /proc/cpuinfo — the ‘flags’ field.
$ cat /proc/cpuinfo | grep flags | grep ida | uniq
If you get output from this command, it tells you that the Linux kernel detected the ‘ida’ feature. ‘ida’ is simply the acronym for Intel Dynamic Acceleration, which is the official name for TurboBoost. I verified that the Envy has this feature turned on.
C states
The Envy BIOS presents cc1/cc6 (advertised as C1/C2) to the OS when the AC adapter is plugged in during boot (call this AC mode) and presents cc1/cc3/cc6 (advertised as C1/C2/C3) if it is running on battery during boot (call this DC mode).
I’ve noticed that the machine idles at around 30C when in DC mode, and idles around 55C in AC mode. This happens in both Windows and Linux, and is likely to be a permanent behavior. So what I’ve been doing is booting the machine unplugged (DC mode) and then after logging in, plugging the power back in. Even after plugging the power in, the machine stays in DC mode, and will idle significantly cooler.
Currently, Linux has some issues with displaying the C state information in AC mode. That’s next on my list of things to investigate, but I just haven’t gotten around to it yet.
Update: (1/20) After some more debugging, we’ve discovered that the BIOS advertises a C2 latency of 245us, which Linux didn’t like. Linux refused to accept latencies > 100us, and that is why the machine idled so hot in AC mode. The fix was to patch Linux and remove the 100us limit for C2. After applying this patch, my machine now idles between 30C and 40C, even in AC mode.
summary
With the most recent BIOS updates, Linux works fairly well on the Envy. The most serious outstanding issue is the screen corruption during suspend/resume, and I hope to get to that after vacation.
If folks out there want to report other issues, drop me a comment and I’ll see what I can do.
Finally, this sucker is hard to photograph because it’s so darn shiny. My first few attempts in using my ghetto-fab lightbox kept on revealing the crappy cardboard interior reflected on the machine’s surface. I figured out that I could use another piece of posterboard to prevent stray reflections of the lightbox and ended up shooting from the side so I didn’t show up in the shot. Still, you can see the reflection of the cardboard in the edge of the lid. Oh well.
This was my setup to get the photo above.
- Posted by alex at 04:39 pm
- Permalink for this entry
- Filed under: geek, photo
- RSS comments feed of this entry
- TrackBack URI
Ooh- the Dr. Dre version, no less! Do you get to keep it for “continued testing”?
Thanks for the Info I’m almost up and running with a custom kernel. Everything seems to be working. The last thing hanging out is the ATI driver. What version if any are you using?
Thanks,
Armin
Nehal, heh, we’ll see.
Armin, what do you mean “hanging out”? I’m just using whatever comes default with Ubuntu Karmic. I haven’t made any changes there.
I downloaded and installed the ati drivers from the ATI website. Just create a debian package using the command seems to be working fairly well. Email me if your curious.
On a side note what kernel version are you running? I applied the latest patches to the current version and it seemed to be working fine.
It will be very interesting once you think you have arrived at a stable solution where Ubuntu 9.10 runs pretty well with F.07 that you write some guide so people can follow the steps to install reliably. Those combinations of 9.10 and additional upstream patches(?) does not seem fully described to us followers.
I got myself an Envy 15 today. It came with BIOS F.04 and still I find the fans run pretty cool. When looking at BIOS update descriptions nothing what so ever seems to have been changed following F.05 that hint at changes related to fan contol, C states, etc. Do you believe such changes exist there although not documented in release notes ?
Gunnar,
My patches were accepted into mainline and are present in Linus’s latest kernel (2.6.33-rc3).
If you are feeling adventurous, you can download it and try it yourself. Otherwise, you will probably have to wait until the next version of Ubuntu to get the changes.
Here’s some help on how to build Linus’s kernel.
https://wiki.ubuntu.com/KernelTeam/GitKernelBuild
Even though the release notes don’t say anything, I still recommend updating to the latest BIOS.
Thanks for you work on this. The C-state issue doesn’t just affect the idle temperature. When you boot the PC with AC power connected, all of the unused cores idle in C1, and for CPU intensive single-threaded process, turbo boost only raises the clock rate to 1.73 GHz or so. But if you boot on DC power and then connect AC, some of the unused cores idle in C6, and turbo boost raises the clock rate to 2.8 GHz. This can be seen with i7z http://code.google.com/p/i7z/, and the performance speed-up is obvious. My code runs 50% faster on the Envy if I first boot on DC power which is a big deal to me.
My configuration: Envy 15, 2.6.32-10 kernel (64 bit), Ubuntu Lucid, F.07 BIOS. X11 problems seemed somewhat reduced for Lucid vs. Karmic, but proprietary ATI driver (fglrx) still doesn’t work.
I tried to build the latest Linus rc kernel last night but ran into a version check problem that I wasn’t able to resolve.
“UTS Release version does not match current version”
Till, thanks for the pointer to i7z. I’ve been extremely busy with the day job recently, but I hope to play with that tool soon.
Gunnar, looks like a problem specific to debian/ubuntu. Can you try asking on #ubuntu? https://help.ubuntu.com/community/InternetRelayChat
Hi again, I was able to build the “2.6.33-rc4″ kernel now on the Envy 15 by applying the patch posted here: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=561569#27 by jean-luc.coulon@wanadoo.fr
Apart from editing in that patch manually everything else is quite straightforward, I followed this kernel build tutorial:
https://wiki.ubuntu.com/KernelTeam/GitKernelBuild
I will test run my build tomorrow.
Regards,
/gunnar (medial.com)
Further note; the build tutorial mentions the make system by default does a debug build.
The turorial hints that when running “make menuconfig” (see tutorial) to go into the “Kernel hacking” settings; then, under “Kernel debugging”, turn OFF “Compile the kernel with debug info”
Sounds like the kernel should run more efficient if not built for debug, is this a reasonable assumption ?
My resulting linux image .deb file size is 33 082 538 bytes.
Very fast, seems like double what I’ve seen on Win7, but fans run too loud for me to move from Win7 so far. Likely due to heat from chips not as well throttled/managed as on Windows 7 ? Heck I have no clue really if there’s any ACPI stuff harnessing the Envy at all in what I built.. I’ll let Alex try his magic.
it’s not clear for me from the article: does the proprietary ATI driver work with the laptop? the ‘only’ issue is the screen corruption after resume?
Gunnar, you’re right that building without debug info will result in (slightly) better performance. However, the tradeoff is that when something does go wrong, you have less information to debug with (duh). It’s a choice you need to make for yourself. I always build with debug turned on since I expect bugs and need info to fix them.
Akos, I’ve never used the proprietary drivers; I’m not very interested in them. With the open-source drivers, there is still screen corruption after resume that I haven’t tracked down yet. I’m still working on C state issues.
For anyone reading this, we’ve discovered the root cause for the C state issue. It was a bug in Linux, not in the BIOS, and we’ve patched it. Machine now idles much cooler!
Linus hasn’t pulled the ACPI tree yet, but I think the fix should appear in 2.6.33-rc5 (when it is released). Alternatively, you can grab the patch from the bugzilla mentioned above and apply it yourself.
Hi, I am running Ubuntu Lucid with 2.6.32.11 kernel. When I suspend and attempt to resume I am getting an entirely blank screen, no sign of the video corruption reported. The laptop does not respond to pings so I am assuming that it is not resumeing correctly. Has anyone else had this problem?
Bart, I had this problem, too on 2.6.32.10. Then I installed the 2.6.33-rc5 kernel which fixes the C-states and thermal and power consumption issues (thanks Alex and others!). I also installed xorg-edgers and updated to the latest xorg stuff and ati driver. With this combination C-states were good and I *could* suspend & resume several times but not hibernate.
Since then things have gone south (no idea what changed) and in the 2.6.33 kernel I can only start X with the vesa driver which doesn’t suspend/resume.
I built rc5 and it indeed runs cooler. It could benefit from some tweaking the fan speed to match that of Windows 7 so I will try to learn this area.
Does it still apply that more C-states are exposed by booting on battery rather than charger connected ?
Gunnar, it is true that you get more C-states when booting in DC.
However, it is not true that the extra C-states improve idle temperature or performance.
Basically, in AC mode, you get “shallow” and “deep” C states. In DC mode, you get the additional “middle” C state.
The “deep” C state saves the most power and allows your machine to idle cool; since you now have “deep” in AC mode, you do not need to boot unplugged.
Before the patch, you only had “shallow” and did not have “deep” at all, which is why the machine ran so hot.
Thanks for all your work Alex. I look forward to the official release, I am not sure if I built the kernel with all params properly. Ubuntu seems to run cooler than Win7 although the right hand fan speed seems slightly higher.
When I tried it today the left side runs very cool (and fan off ?) while the right hand palmrest gets a bit warm but overall much less than before.
Hi,
I’ve been running fedora 12 of a usb drive for a month or so. I was wondering what the position was regarding the bios and the esata connection.
Gunnar,
For the problem “UTS Release version does not match current version”, I was able to download the latest kernel-package from debian:
kernel-package_12.032_all.deb
And the kernel built fine for me without any errors. Otherwise I followed the instructions placed here for kernel linux-2.6.33-rc7.
https://wiki.ubuntu.com/KernelTeam/GitKernelBuild
Hi, thank you for all your work! I just received my new envy 15 and it is a great machine. I installed Ubunutu 9.10 64 via wubi (to avoid messing with the partitions), and the wireless does not work. Network manager is greyed out and says “device not ready”. I can turn it on and off via the hotkey and I get messages via dmesg but I can never get wireless to fully enable and let me connect. It works well in Windows and I’ve tried everything in the Ubuntu wireless guides. Anyone else had any problems like this, any suggestions? Thanks!
I should also note that: “sudo lshw -C network” shows the adapter fine and appears to have the driver loaded, etc, but says “DISABLED” next to the device. Hitting the wirless hotkey doesn’t seem to work as it’s still disabled.
Keep up the great work Alex, I’m looking forward to running linux on an Envy13 one day soon..
Lanson, try grabbing the linux-backports-modules-karmic-generic package to get the wireless connectivity established.
IDA is NOT the official name for Intel Turbo Boost.
Many thanks Scott! Installing the suggested module worked like a charm! I can’t help but wonder why an hour long google search and the Ubuntu docs didn’t ever mention anything about that… Oh well, thanks for your help, I am up and running!
Folks, I’ve gotten the C2 fix backported into Ubuntu Karmic, so the need to run a custom kernel should mostly go away.
After talking with the Ubuntu kernel team, the ACPI namespace error fix will not go into Karmic unfortunately, since it is a large change and they do not want to disturb a mostly stable codebase. The good news is that the error is actually not that big a deal, mostly talking about the ambient light sensor, which isn’t terribly important for functionality.
I will work on getting the ACPI fix into Lucid. Still need to work with the upstream ACPI maintainer on that one.
I haven’t made much progress on the screen corruption issue, but I hope to get there soonish.
Hi Alex,
Thanks for this blog entry (more like a website, now :-), dedicated to Linux on the HP Envy 15. I just got my hands on one, and have been trying to get Debian/unstable to work. The only out-of-distribution thing I have currently is the kernel (Linux 2.6.33-rc7).
Still, the most annoying thing for me is X. I’ve seen you posted a screenshot here:
http://chizang.net/alex/tmp/corrupt-radeon2.jpg
Well, my screen sometimes (not all times) looks like that *without* having to suspend/resume. Also, an easy trigger is using ‘xrandr’ to modify the resolution. For example, if I do:
xrandr -s 1680×1050
xrandr -s 1920×1080
X becomes corrupted, with “square”-ish patterns. Since this happens quite frequently with KDE pop-ups and messages, it can be quite tricky to use the machine.
So, curiously, my symptoms are a bit different than yours. I’m currently using the radeon driver from Debian experimental: 1:6.12.99+git20100201.a887818f-1
Which appears to be the Git version of Feb/2nd. Which xorg driver and version are you using?
Thanks,
Livio
I just got my HP Envy 15, and I can’t install Ubuntu on it for the very simple reason that the boost CD won’t boot from the external USB DVD drive. I’m sent back to initramfs, with a message saying it cannot find a live filesystem. I’ve tried both 9.10 and 9.04, with the same results.
I also tried to install the additional windows boot option via wubi.exe, but actually I get the very same result here as well..
you guys don’t even mention this being a hurdle – am I missing something totally obvious?
Hi Akos,
To be honest, I did not want to burn a CD to install Debian (seems like a waste to burn a CD to use only once). So I used a USB memory stick to install Debian. However, the images available from Debian use a kernel version (2.6.30), which did not fully support either the wifi _or_ the wired network card. Since I was doing a net-install, this was a problem.
So I downloaded a non-official image from here:
http://kmuto.jp/debian/d-i/
( blog at http://kmuto.jp/b.cgi )
The 2.6.32 image worked out great. It booted fine, and the WiFi worked great. I installed the amd64 version, but assume the i386 would also work fine.
For Ubuntu, there is a dedicated website for installation from a USB stick:
https://help.ubuntu.com/community/Installation/FromUSBStick
Hopefully you have a USB stick lying around, like I did. And hopefully Ubuntu images are more up-to-date then Debian ones.
Good luck.
Alex,
You referenced your machine idling around 30C after getting the C-states to register properly. I’ve tried the 2.6.33-rc7 and 2.6.32-13.19 from Lucid… both are reporting 2/3 states for AC/DC and both idle around 43-47C. That in itself wouldn’t be that big of a deal, but the machine doesn’t seem to want to fully rest on a suspend. The fan on the right (GPU?) is always running high, and the machine runs warm even in a “suspended” state. Are you seeing this? Any thoughts? The suspend is handled in Windows fine.
Possibly related: the machine has trouble powering itself off on a shutdown. Everything seems to spin down (except the fan on the right), but the power is still active.
Akos: try the USB/eSata port (the port on the right, closest to the front of the machine). I have seen inconsistent results booting from the dedicated USB 3.0 slots if you have one of the more recent machines. I also recommend using a USB stick.
I solved the booting / install issue since then by applying a trick: have the machine boot from the DVD, but include the install image on the SD Card reader. the installer finds the install files there fine..
the grub in 9.10 didn’t recognize the internal RAID controller, so it wasn’t willing to set up properly. going for 10.04 alpha solved this problem for me.
see my experiences in detail here: http://ubuntuforums.org/showthread.php?p=8824792
I wonder about the heat in this machine.. Alex states above that his patch solves the power state problem, and that his laptop idles at around 30 C. he also says that his patch was incorporate into 2.6.33-rc3.
I booted the system with 2.6.33-rc8, but for me, the laptop idles at around 49-50 C. am I missing something? how can I check that Alex’s patch is in there, and working?
Livio, just posted an update for the screen corruption problem.
Scott, sorry for misleading you; 30C idle is very cool for me too, and I only saw that maybe once, leaving the machine alone for an hour on a cool glass tabletop. Much more common is high 30s, low 40s (C) idle. As far as successfully suspending and resuming, I’ve noticed that upstream doesn’t handle that very well. I’ve only gotten S/R to work with the modified Karmic kernel. Hopefully, after the ACPI C2 patch hits Karmic, you can just use the default kernel and have a working S/R.
Akos, you wrote that you have the 2nd generation Envy, which I unfortunately do not have. However, 33-rc8 should definitely have the ACPI C2 patch, so I’m not sure why it’s idling so hot. Can you email me the output of “grep . /proc/acpi/processor/CPU0/*” ? Thanks.
Alex,
sure, what is your e-mail address?
as for going for 2.6.33-rc8, I’d need to patch the kernel with the clickpad patch – how do you rebuild a latest mainline kernel? I tried to rebuild from the Ubuntu mainline PPA, but the source repository doesn’t work, and can’t find the .dsc file for the packages anywere :(
Akos
Hi Alex,
Thank you for posting the Xorg workaround. I added the “EXANoDownloadFromScreen” “true” option to my xorg.conf and now all is well. No more corrupted KDE pop-ups, and fonts on my terminals.
Suspend/resume also works well. I even get fancy composite effects on KDE.
Awesome. :-)
for me, suspend doesn’t work either, see the details here:
https://bugs.launchpad.net/ubuntu/+bug/522263
I wonder how it works for you guys…
Does your HP envy 15 have a USB 3.0 host on it? I think if it has a core i7, it has USB3. I’ve been getting reports of my xHCI driver failing on the envy 15, but I don’t have a laptop myself to test, and my reporters seem to have stopped responding.
Hi HP Envy Linux users,
Just a quick comment about the X screen corruption. Debian/experimental just released version 6.12.191 of the ATI/Radeon xorg driver, and I tried it out. With the new driver, I no longer need to disable the EXA Download from screen option. I don’t get corrupted pop-ups in KDE, as before. Also suspend/resume seems to work fine, and X shows no corruption (so far, I’ve been testing for a couple of hours).
The reason why I mention this is that there is potential for performance degradation when disabling the DownloadFromScreen EXA option. Plus, it’s always more pleasant to use software when it “just works” out of the box.
Cheers,
Livio
Anyone else having trouble with shutdown and suspend/resume with fglrx?
It seems most of you are using the OSS driver but I have the 5830 which isn’t well supported yet.
When I try to shutdown it just hangs.
This has happened with my Arch Linux and Ubuntu installs. I’ve tried kernel 2.6.31, 2.6.32, and 2.6.33.
@ babyhuey
I have the same problem here, and same card. I think the system halts okay, but it just does not fully power off. I have to hold the power button down to get it to fully turn off. I think it is safe to do this if you give it 10 seconds or so…
@ Lanson
Have you found any errors in your logs that are relevant?
I assume still you need to boot the Envy 15 with charger unattached in order for it to expose all C states which leads to a cooler operation. Once booted the charger can be connected.
I’ve been using ULV processor machines for a while now so I’m spolied with the silent operation there and 8/10 hours battery.
But once I get the fan speed lowered a bit on Ubuntu (to WIn7 levels) the Envy 15 will be hard to resist there. It seems there’s headroom for some passive cooling now that Alex made it so low temperature.
Rather a bit warmer and more silent is what I would want.
I have not seen anything in the logs (syslog, dmesg), so I am wondering if the logging processes are stopped correctly by that point and the system is indeed halted but just can’t be fully powered off. I haven’t tried suspend / resume, but my guess is it won’t work.
@ Lanson
Yeah, my thoughts are the same with the logs.
Suspend doesn’t work either and again nothing in the logs.
The only meaningful difference between my Envy 15 and those that have working shutdown/suspend is the video card and the CPU.
In terms of the video card, I’ve tried the OSS drivers (with and without KMS), fglrx, and vesa. All give the same results.
Has anyone tried running Live CD on the most recent bios (F.18)? I get “Fail in evaluating the _REG object of EC device. Broken bios is suspected….”. I tried F17 and F19 as well and get the exact same error!
Excellent guide, thank you all!
This is the definite source for information about Linux on the Envy 15″. There are others, but this is definitely the best.
Each time I thought of something to ask, I just went back and read your comments again. Even the problem with compiling the kernel due to mismatch in UTS_RELEASE is described.
So now I’m running my own patched build of the 2.6.33 kernel (last time I built a kernel was in ’95 or thereabouts).
Everything seems to work except bluetooth. Does anybody know how to configure bluetooth in the kernel, or even which chip is in there?
Follow up from my last post.
Now BT works too. The only thing I did was to go into windows 7 and enable the BT chip, it had not been started before…
I don’t know if it was HW enabled by some voodoo in windows?
It could also be that I have tried to cut down the number of options in the kernel to a minimum. Maybe I removed something that was conflicting with BT.
So, the last thing remaining is suspend / resume. Which I hear is really difficult to get to work in linux on any laptop.
Once again. Thanks people!
Hi all,
Just got my Envy today, running Ubuntu off a USB stick and would love to resolve the mouse issues. I noticed that the only mention for a fix is of the patch by Takashi Awai; is that what everyone is using?
I’m totally new to Linux, so I’m not sure exactly how I would apply the patch, but my guess is something like:
1. Download the source for the latest kernel from Git, as per the link posted above.
2. Locate the synaptics.c source and apply the patch <– not sure how to do this
3. Finish building the kernel, as per the tutorial.
Am I on the right track? Leaving anything out? Doing it totally wrong?
Thanks,
Dan
Hi,
Alex, thanks so much for this blog /guide! The screen corruption thing was driving me nuts!
Anyway, I seem to be the only one who HAS NO SOUND. I am sort of inclined to think it’s because I did something stupid, since if it were a bug, everyone would have it. I installed alsamixer, and no channels are muted, but there is just no output coming from any channels, even though various media players (tried a few) claim to be playing a track. BTW, if I plug in a headphone, I can hear very faint and distorted audio. Don’t know if that info helps.
Anyway, any help would be appreciated.
–Tristan
Tristan,
I have an Envy 15 and have the same problem with sound. The light on the “sound off” key (aka f11) is always on, though toggling the key and the “sound up / sound down” keys (aka f9 / f10) shows a sound icon indicating sound was turning off and on and volume being adjusted upwards and downwards.
I tried the edit to /etc/modules.d/alsa-base.conf per
http://ubuntuforums.org/showthread.php?t=1304990
with no luck (though that post was dated 2007 so devices have probably been updated since). The following shows nothing has taken control from pulseaudio
sudo fuser -v /dev/dsp* /dev/snd* /dev/seq*
Alsamixer shows all channels on max. System -> Preferences -> Sound -> Applications shows npviewer.bin registering usage by an application.
Any advice would be much appreciated. It would be a super laptop if I could just get sound working. Otherwise, back to Windows (sigh).
Opened System -> Preferences -> Sound, then on the Output tab switched from “R700 Audio Device [ Radeon HD 4000 Series] Digital Stereo (HDMI)” to “Internal Audio Analog Stereo” and following a reboot, sound was there. Not sure if it was the switch or the reboot or a combination (or perhaps even one of the other things I tried in combination with the reboot), but sound works just fine now.
Curiously, the “sound off” key is still lit even though the sound works.
Starky,
I guess you are using gnome, and so our options for sound are arranged a little differently. (I’m on KDE)
Can’t do what you did cuz’ my devices are named differently, but I’ll try to do something analogous a little later today.
I just wanted to let you know that I plugged in some USB headphones, and I have sound from them. (I just tried “testing” them in the device settings panel, I haven’t yet gotten any output from applications, though).
Don’t know what this might mean in terms of why my speakers don’t work (yes, they work in windows :) ), but I thought I would post my experience.
Again, anyone with any insight would be very much appreciated.
–Tristan
Starky,
I looked at the device configuration panel in my KDE desktop (System Settings –> Multimedia), which I think is analogous to System -> Preferences -> Sound in Gnome. I didn’t have any luck though. Your problem, and the solution you found makes sense, though. Your output was set to go to the HDMI device, which is the digital output that would usually go to a HDTV or a monitor. That output is separate from the internal speakers. “Internal Audio Analog Stereo”, I would guess, is the name of your internal speakers, so when you changed to this output, you got sound. Yay!
I, however, am still stuck, although I have used my USB headphones to connect to some external speakers, so this works okay in the (hopefully short) interim. I can listen to Tom Waits while writing this comment at least without being tethered by a headphone cable.
-Tristan
For those of you having problems with sound, do you have problems in other kernels? For me, sound works perfectly in 2.6.31.19, but not since I upgraded to 2.6.31.20. So, I can still just boot to the older kernel and it works. Seems to be a problem for others as well (https://answers.launchpad.net/ubuntu/+source/alsa-driver/+question/103219), maybe the ALSA modules weren’t compiled for 2.6.31.20?
Has anyone else experienced any issues with the Clickpad patch? I installed it under Lucid in the 2.6.34-rc3 kernel and can see “Clickpad mode enabled” in dmesg, but the mouse behavior hasn’t changed at all.
For those with 2nd gen Envy 15s that have the shutdown/suspend issue, there is a bug report created by Akos.maroy @ https://bugs.launchpad.net/ubuntu/+source/linux/+bug/557072
Hi guys I am looking to getting the HP Evny 14 (2rd gen) I don’t know if its comes with the same Hardware as Evny 15 has anyone tried Ubuntu on Hp Envy 14. also I would like to know if the multi touch touch pad works on ubuntu
[...] [...]
Anyone already use the enable/disable led from touchpad??
I’ve discovered a problem with the latest 2.6.34 (and .33) kernel on my HP Envy 15t:
https://bugzilla.redhat.com/show_bug.cgi?id=628122
I reported it on Fedora, but the problem is present in Ubuntu as well.
Hi all, as being new to Linux I’ve found this site while searching for solutions. My current situation is this: in attempts to install Ubuntu 10.10 and 11.04 upon my Hp Envy 17 3D (i7-2630QM), the install “seems” to go fine then however upon re-boot I consistently receive a thermal shutdown error sometimes in combination with the fans spinning rapidly prior to the aforementioned error prompt. Sadly due to this I’ve not been able to enjoy the fruit which is Linux/Ubuntu.
Thanks in advance for any guidance/assistance.
I have been having the exact problem as Jamari I have HP Envy 17″ 3D and i have it dual booted with win7 and Ubuntu 11.04 Windows works fine be if i log in to Ubuntu and reboot i get the Thermal shutdown error even though the machine doesn’t seem to be over heating