Skip to contentSkip to conversation

xen/mtrr: Add mtrr_if support for Xen mtrr

PATCH Sander Eikelenboom 1 participant 1 message Sep 28, 2010

lore
Share & tools
lore
#0 lore raw
Quoted 25 lines · * stefano.stabellini@xxxxxxxxxxxxx <stefano.stabellini@xxxxxxxxxxxxx> wrote:
Quoted 9 lines

From: Stephen Tweedie <sct@xxxxxxxxxx>

Add a Xen mtrr type, and reorganise mtrr initialisation slightly to allow the mtrr driver to set up num_var_ranges (Xen needs to do this by querying the hypervisor itself.)

[ Impact: add basic MTRR support ]

Quoted 8 lines

---
arch/x86/kernel/cpu/mtrr/Makefile | 2 +- arch/x86/kernel/cpu/mtrr/main.c | 3 + arch/x86/kernel/cpu/mtrr/mtrr.h | 7 ++ arch/x86/kernel/cpu/mtrr/xen.c | 110 +++++++++++++++++++++++++++++++++++++

4 files changed, 121 insertions(+), 1 deletions(-)

create mode 100644 arch/x86/kernel/cpu/mtrr/xen.c

Still NAK, for the very same reasons as we NAK-ed it the previous time:
/proc/mtrr is a problematic and complicated legacy interface that should die. Any modern X server will do the right thing via PAT.

Quoted 7 lines

Also, please get the Ack of at least one x86 maintainer for x86 patches.

Thanks,

Ingo
--

I can't find the MTRR API to be officially deprecated with any schedule for removal (at least i couldn't find it in feature-removal-schedule.txt).
KVM has had patches for MTRR as well, if it's so deprecated .. why has it ? .. legacy support for a not officially deprecated API perhaps ? To support CPU's that support MTRR but not PAT ?

If you think the MTRR interface should die, and should die fast, why hasn't it been added to feature-removal-schedule.txt for a complete removal in a not so distant future, since this is the only way to force all users to switch to PAT.

Perhaps it would be fair to ask for a working PAT implementation besides MTRR, but since it is not officially deprecated completely objecting a MTRR one just for Xen seems a bit unfair to me.

--
Sander