emacs running slow on Fedora 16 and nvidia

asked 2011-11-17 23:35:25 -0600

espeed
1 ●1 ●1 ●2

updated 2011-11-18 00:01:39 -0600

I upgraded from Fedora 14 to Fedora 16, and emacs is running painfully slow.

Emacs is slow to open, and there is a lag when moving the cursor position, opening files, and searching within frames.

Fedora 14 was fine, and all the hardware is the same. This is with a 3.1 kernel with the latest nvidia drivers on a three-screen display.

I switched to XFCE thinking Gnome 3 was the culprit, but the lag remains.

Are there issues with Fedora 16, the 3.1 kernel, and/or nvidia drivers on multihead displays? Or where else should I be looking?

$ uname -a
Linux love 3.1.1-1.fc16.x86_64 #1 SMP Fri Nov 11 21:47:56 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux

$ yum info akmod*
Loaded plugins: langpacks, presto, refresh-packagekit
Installed Packages
Name        : akmod-nvidia
Arch        : x86_64
Epoch       : 1
Version     : 290.06
Release     : 1.fc16.2
Size        : 8.7 M
Repo        : installed
From repo   : rpmfusion-nonfree-updates
Summary     : Akmod package for nvidia kernel module(s)
URL         : http://www.nvidia.com/
License     : Redistributable, no modification permitted
Description : This package provides the akmod package for the nvidia kernel modules.

Name        : akmods
Arch        : noarch
Version     : 0.3.7
Release     : 1.fc16
Size        : 27 k
Repo        : installed
From repo   : rpmfusion-free
Summary     : Automatic kmods build and install tool
URL         : http://rpmfusion.org/Packaging/KernelModules/Akmods
License     : MIT
Description : Akmods startup script will rebuild akmod packages during system
            : boot while its background daemon will build them for kernels right
            : after they were installed.

edit retag flag offensive close merge delete

Comments

Emacs IS slow, try Vim instead :-)

lzap ( 2011-11-18 06:16:53 -0600 )edit

add a comment

1

answered 2011-12-23 19:03:33 -0600

Ved Vyas
71 ●1 ●3 ●2

I had the same problem, though it affected emacs -nw as well (didn't try emacs-nox). Also, using nouveau not the blob. Anyway, that turned out to be orthogonal to the actual issue in my case. For a possible solution, skip to the end. I did:

ltrace emacs

and stopped it at when Emacs "hung" to see what library calls were being made. The last chunk of relevant output was:

access("/usr/share/emacs/23.3/etc/charse"..., 0) = 0
gethostname("MYHOSTNAME", 255)                    = 0
index("MYHOSTNAME", '.')                          = NULL
getaddrinfo("MYHOSTNAME", NULL, 0x7fffab63a9d0, 0x7fffab63aa00 <unfinished ...>
(3, 0x3203823260, 2054, -1, 0)                   = 0x3203823260
(0, 0, 0, 3, 0x7f2f6d1f6040)                     = 0x3203823260
(3, 0x3203823260, 2054, -1, 0)                   = 0x3203823260
(0, 0, 0, 3, 0x7f2f6cff34a8)                     = 0x3203823260
(3, 0x3203823260, 2054, -1, 0)                   = 0x3203823260
(0, 0, 0, 4, 0x7f2f6cdeb718)                     = 0x3203823260

where MYHOSTNAME represents my true hostname. The last call to getaddrinfo suggested that the pause was due to a delay in resolving my hostname to an IP, or something like that. At this point, there could be a couple of solutions, some more correct than others.

What I did:

I pinged my hostname, and found that it was resolving to the IP on my local network after a similarly long pause. So, I just added my hostname to the 127.0.0.1 line in /etc/hosts. Et voilà: Emacs started up quickly. Well as quickly as your init file allows :)

To any commenters reading this: I would be interested if there's a better solution.