[Pdbg] [PATCH] Revert "htm: Use splice() to copy dump"

Rashmica Gupta rashmica.g at gmail.com
Fri Dec 4 08:42:27 AEDT 2020


On Tue, 2020-12-01 at 12:25 +1100, Michael Neuling wrote:
> On Tue, 2020-12-01 at 11:52 +1100, Rashmica Gupta wrote:
> > On Tue, 2020-12-01 at 11:37 +1100, Michael Neuling wrote:
> > > On Thu, 2020-11-26 at 13:38 +1100, Rashmica Gupta wrote:
> > > > This reverts commit 436eb8c74fb4a762b61837ee27ddbd6b5fe21334.
> > > > Unable to use splice on newer kernels due to 36e2c7421f02 ("fs:
> > > > don't
> > > > allow splice read/write without explicit ops"). So revert back
> > > > to
> > > > plain
> > > > old read/write.
> > > 
> > > Is there a performance penalty with this and if so how much? 
> > 
> > On older kernels most probably, but does it matter when splice just
> > fails on newer kernels? Do you want to dectect kernel version and
> > choose whether to use splice or not?
> 
> (adding people back on CC)
> 
> If the performance hit is particularly bad, then we might want the
> splice()
> optimisation, but if it's only a small hit then I wouldn't bother.

With my eyeball statistics, there is no noticable difference with
dumping a 1.7GB file.
> 
> Mikey
> 



More information about the Pdbg mailing list