Hi again,
updating the table after Yves-Alexis' comment on PCID. Rerunning the test
with -cpu=Haswell to enable PCID gave me much better numbers :
On Sun, Jan 07, 2018 at 11:18:56AM +0100, Willy Tarreau wrote:
> Hi,
>
> I managed to take a bit of time to run some more tests on PTI both
> native
Hi again,
updating the table after Yves-Alexis' comment on PCID. Rerunning the test
with -cpu=Haswell to enable PCID gave me much better numbers :
On Sun, Jan 07, 2018 at 11:18:56AM +0100, Willy Tarreau wrote:
> Hi,
>
> I managed to take a bit of time to run some more tests on PTI both
> native
On Mon, Jan 08, 2018 at 09:26:10PM +0100, Yves-Alexis Perez wrote:
> On Mon, 2018-01-08 at 19:26 +0100, Willy Tarreau wrote:
> > You're totally right, I discovered during my later developments that
> > indeed PCID is not exposed there. So we take the hit of a full TLB
> > flush twice per syscall.
On Mon, Jan 08, 2018 at 09:26:10PM +0100, Yves-Alexis Perez wrote:
> On Mon, 2018-01-08 at 19:26 +0100, Willy Tarreau wrote:
> > You're totally right, I discovered during my later developments that
> > indeed PCID is not exposed there. So we take the hit of a full TLB
> > flush twice per syscall.
On Mon, 2018-01-08 at 19:26 +0100, Willy Tarreau wrote:
> You're totally right, I discovered during my later developments that
> indeed PCID is not exposed there. So we take the hit of a full TLB
> flush twice per syscall.
So I really think it might make sense to redo the tests with PCID, because
On Mon, 2018-01-08 at 19:26 +0100, Willy Tarreau wrote:
> You're totally right, I discovered during my later developments that
> indeed PCID is not exposed there. So we take the hit of a full TLB
> flush twice per syscall.
So I really think it might make sense to redo the tests with PCID, because
Hi Yves-Alexis,
On Mon, Jan 08, 2018 at 06:07:54PM +0100, Yves-Alexis Perez wrote:
> On Sun, 2018-01-07 at 11:18 +0100, Willy Tarreau wrote:
> > - the highest performance impact on VMs comes from having PTI on the
> > guest kernel (-45%). At this point it makes no difference whether
> >
Hi Yves-Alexis,
On Mon, Jan 08, 2018 at 06:07:54PM +0100, Yves-Alexis Perez wrote:
> On Sun, 2018-01-07 at 11:18 +0100, Willy Tarreau wrote:
> > - the highest performance impact on VMs comes from having PTI on the
> > guest kernel (-45%). At this point it makes no difference whether
> >
From: Willy Tarreau
> Sent: 07 January 2018 10:19
...
> The impact inside VMs is quite big but it's not where we usuall install
> processes sensitive to syscall performance. I could find an even higher
> impact on a packet generation program dropping from 2.5 Mpps to 600kpps
> in the VM after the
From: Willy Tarreau
> Sent: 07 January 2018 10:19
...
> The impact inside VMs is quite big but it's not where we usuall install
> processes sensitive to syscall performance. I could find an even higher
> impact on a packet generation program dropping from 2.5 Mpps to 600kpps
> in the VM after the
On Mon, 2018-01-08 at 18:07 +0100, Yves-Alexis Perez wrote:
> On Sun, 2018-01-07 at 11:18 +0100, Willy Tarreau wrote:
> > - the highest performance impact on VMs comes from having PTI on the
> > guest kernel (-45%). At this point it makes no difference whether
> > the host kernel has it
On Mon, 2018-01-08 at 18:07 +0100, Yves-Alexis Perez wrote:
> On Sun, 2018-01-07 at 11:18 +0100, Willy Tarreau wrote:
> > - the highest performance impact on VMs comes from having PTI on the
> > guest kernel (-45%). At this point it makes no difference whether
> > the host kernel has it
On Sun, 2018-01-07 at 11:18 +0100, Willy Tarreau wrote:
> - the highest performance impact on VMs comes from having PTI on the
> guest kernel (-45%). At this point it makes no difference whether
> the host kernel has it or not.
Hi Willy,
out of curiosity, is the pcid/invpcid flags
On Sun, 2018-01-07 at 11:18 +0100, Willy Tarreau wrote:
> - the highest performance impact on VMs comes from having PTI on the
> guest kernel (-45%). At this point it makes no difference whether
> the host kernel has it or not.
Hi Willy,
out of curiosity, is the pcid/invpcid flags
Hi,
I managed to take a bit of time to run some more tests on PTI both
native and hosted in KVM, on stable versions built with
CONFIG_PAGE_TABLE_ISOLATION=y. Here it's 4.9.75, used both on the
host and the VM. I could compare pti=on/off both in the host and the
VM. A single CPU was exposed in the
Hi,
I managed to take a bit of time to run some more tests on PTI both
native and hosted in KVM, on stable versions built with
CONFIG_PAGE_TABLE_ISOLATION=y. Here it's 4.9.75, used both on the
host and the VM. I could compare pti=on/off both in the host and the
VM. A single CPU was exposed in the
16 matches
Mail list logo