Re: size of kernel after gcc4.2 upgrade

Quoting Ruslan Ermilov <ru@xxxxxxxxxxx> (from Fri, 25 May 2007 14:52:03 +0400):

On Fri, May 25, 2007 at 11:51:46AM +0200, Roman Divacky wrote:
hi

I just noticed that
spravci ~# ls -l /boot/kernel*/kernel
-r-xr-xr-x 1 root wheel 3389971 May 18 12:03 /boot/kernel.old/kernel*
-r-xr-xr-x 1 root wheel 5211345 May 25 11:47 /boot/kernel/kernel*

ie. after gcc42 import the kernel size increased roughly by 60% thats
a little too much. is there any forgotten option or something that
makes the kernel grow so big?

My numbers are different (1% for the GENERIC kernel, and this
is with all code changes in-between):

-r-xr-xr-x 1 root wheel 8942239 Apr 11 16:20 kernel.old/kernel
-r-xr-xr-x 1 root wheel 9035021 May 24 16:23 kernel/kernel

bde@ also reported a 1% code bloat in another thread, so I think
you should look at your own changes.

My minimal kernel without additional changes shows:
-r-xr-xr-x 1 root wheel 2.5M 7 Apr 17:22 /boot/kernel.old/kernel*
-r-xr-xr-x 1 root wheel 8.1M 7 Apr 17:22 /boot/kernel.old/kernel.symbols*
-r-xr-xr-x 1 root wheel 3.9M 24 Mai 18:46 /boot/kernel/kernel*
-r-xr-xr-x 1 root wheel 14M 24 Mai 18:46 /boot/kernel/kernel.symbols*

But this is with debugging (-g) enabled. How's the .symbols generated? Can there be some new stuff in the kernel which needs to be handled in the .symbols file generation?

Bye,
Alexander.

--
The relative importance of files depends on their cost
in terms of the human effort needed to regenerate them.
-- T. A. Dolotta

http://www.Leidinger.net Alexander @ Leidinger.net: PGP ID = B0063FE7
http://www.FreeBSD.org netchild @ FreeBSD.org : PGP ID = 72077137
_______________________________________________
freebsd-current@xxxxxxxxxxx mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscribe@xxxxxxxxxxx"

Relevant Pages

  • Re: Does "c" loose some velocity after leaving a dense medium back into space?
    ... John C. Polasek wrote in ... > Espace needed to solve the ultraviolet problem. ... The wavelength associated with this frequency is 1.1e-13m which is much ...
    (sci.physics)
  • Re: Does "c" loose some velocity after leaving a dense medium back into space?
    ... John C. Polasek wrote in ... > Espace needed to solve the ultraviolet problem. ... The wavelength associated with this frequency is 1.1e-13m which is much ...
    (sci.physics.relativity)
  • Re: 2.6.23 kernel
    ... Thats correct Scott RH4, sorry but once bitten twice shy, I am very happy not using vendor kernels as they tend to be too bloated anyway, core servers only need small kernels, enough to run what they need to, no need to have built in network crud i'll never use, no need to have 30K modules inplace that will never be used, vendors do this so its a plug and play type OS, thats fine for a desktop, not fine for a server. ... I've never had them rendered unbootable by a vendor kernel ...
    (Fedora)
  • Re: Real-Time Preemption, -RT-2.6.10-rc2-mm3-V0.7.31-13
    ... >Gene Heskett wrote: ... snip, but thats it. ... AND I just discovered that I am not running the -13 kernel, ... and this is true whether the fonts are compiled ...
    (Linux-Kernel)
  • Re: [kde] Suspend Issues, or soft kernel locks + no networking, which is worse?
    ... And thats kernel 3.0 essentially) ... The Fedora devs were worried about backwards compatibility ... with external kernel modules and other applications that depend on ... And yes, thats true. ...
    (KDE)