Rusty Russell wrote:
> From: Mark McLoughlin <[EMAIL PROTECTED]>
>
> hdr->csum_start is the offset from the start of the ethernet
> header to the transport layer checksum field. skb->csum_start
> is the offset from skb->head.
>
> skb_partial_csum_set() assumes that skb->data points to the
> ether
On Tuesday 10 June 2008 17:57, Jeremy Fitzhardinge wrote:
> Nick Piggin wrote:
> > On Tuesday 10 June 2008 17:35, Isaku Yamahata wrote:
> >> This patch is ported one from 534:77db69c38249 of linux-2.6.18-xen.hg.
> >> Use wmb instead of rmb to enforce ordering between
> >> evtchn_upcall_pending and
On Tuesday 10 June 2008 17:35, Isaku Yamahata wrote:
> This patch is ported one from 534:77db69c38249 of linux-2.6.18-xen.hg.
> Use wmb instead of rmb to enforce ordering between
> evtchn_upcall_pending and evtchn_pending_sel stores
> in xen_evtchn_do_upcall().
There are a whole load of places in
If we're not gonna do anything (case in which failure is already
reported), we do not need to even bother with calculating the linear rip.
This is a nitpick, but I saw it while doing some testing, so here's
the patch.
Signed-off-by: Glauber Costa <[EMAIL PROTECTED]>
---
arch/x86/kvm/x86.c |4
define xen paravirtualized instructions for hand written assembly code.
Signed-off-by: Yaozu (Eddie) Dong <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
include/asm-ia64/xen/inst.h | 450 +++
1 files changed, 450 insertions(+), 0
implement xen pv_iosapic_ops for xen paravirtualized
iosapic.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/xen/xen_pv_ops.c | 52
1 files changed, 52 insertions(+), 0 deletions(-)
diff --git a/arch/ia64/xen/xen_pv_ops.c b/arch/ia6
paravirtualize entry.S for ia64/xen by multi compile.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/xen/Makefile |2 +-
arch/ia64/xen/xen_pv_ops.c | 18 ++
include/asm-ia64/xen/inst.h |8
3 files changed, 27 insertions(+), 1 deletions(-)
define xen machine vector for domU.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/Makefile |2 ++
arch/ia64/kernel/acpi.c|5 +
arch/ia64/xen/Makefile |2 ++
arch/ia64/xen/machvec.c|4
include/asm-ia64/machvec.h |
define include/asm-ia64/xen/irq.h to define the number of
irqs which xen needs.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/kernel/nr-irqs.c |1 +
include/asm-ia64/xen/irq.h | 44
2 files changed, 45 insertions(+), 0 deletions
This patch implements xen version of pv_init_ops to do various
xen initialization.
This patch includes ia64 counter part of
9e124fe16ff24746d6de5a2ad685266d7bce0e08 and
83abc70a4c6e306f4c1672e25884322f797e4fcb for xen early printk support.
Signed-off-by: Akio Takebe <[EMAIL PROTECTED]>
Signed-off-
preliminary support for save/restore.
Although Save/restore isn't fully working yet, this patch is necessary
to compile.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/xen/Makefile |2 +-
arch/ia64/xen/{time.h => suspend.c} | 40
paravirtualize DO_SAVE_MIN in minstate.h for xen.
Signed-off-by: Yaozu (Eddie) Dong <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
include/asm-ia64/xen/inst.h |2 +
include/asm-ia64/xen/minstate.h | 134 +++
2 files changed,
implement xen pv_time_ops to account steal time.
Cc: Jeremy Fitzhardinge <[EMAIL PROTECTED]>
Signed-off-by: Alex Williamson <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/xen/Makefile |2 +-
arch/ia64/xen/time.c | 165 +++
Xen implements grant tables which is for sharing pages with
guest domains.
This patch implements arch specific part of grant table initialization.
and xen_alloc_vm_area()/xen_free_vm_area() which are helper functions
for xen grant table.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/
introduce CONFIG_PARAVIRT_GUEST, CONFIG_PARAVIRT for
paravirtualized guest.
introduce CONFING_XEN, CONFIG_IA64_XEN for xen.
Signed-off-by: Alex Williamson <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/Kconfig | 32
arch/
define xen pv_cpu_ops which implementes xen paravirtualized
privileged instructions.
Signed-off-by: Yaozu (Eddie) Dong <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/xen/xen_pv_ops.c | 114
1 files changed, 114 ins
paravirtualize ivt.S for xen by multi compile.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/xen/Makefile | 16 -
arch/ia64/xen/xenivt.S | 52 +++
include/asm-ia64/xen/inst.h |1 +
3 files changed, 68 insertio
Recipe for useing xen/ia64 with pv_ops domU.
Signed-off-by: Akio Takebe <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
Documentation/ia64/xen.txt | 183
1 files changed, 183 insertions(+), 0 deletions(-)
create mode 100644
On ia64/xen, pointer arguments for hypercall is passed
by pseudo physical address(guest physical address.)
So such hypercalls needs address conversion functions.
This patch implements concreate conversion functions for
such hypercalls.
Signed-off-by: Akio Takebe <[EMAIL PROTECTED]>
Signed-off-by:
This patch enables elf note based xen startup for IA-64, which gives the
kernel an early hint for running on xen like x86 case.
In order to avoid the multi entry point, presumably extending booting
protocol(i.e. extending struct ia64_boot_param) would be necessary.
It probably means that elilo also
Xen paravirtualizes interrupt as event channel.
This patch defines arch specific part of xen event channel.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
include/asm-ia64/xen/events.h | 50 +
1 files changed, 50 insertions(+), 0 deletions(-)
crea
On ia64/xen, pointer argument for the hypercall is passed
by pseudo physical address (guest phsysical address.)
So it is necessary to convert virtual address into pseudo physical
address rigth before issuing hypercall. The frame work is called
xencomm. This patch implements arch specific part.
Sig
Nick Piggin wrote:
> On Tuesday 10 June 2008 17:35, Isaku Yamahata wrote:
>
>> This patch is ported one from 534:77db69c38249 of linux-2.6.18-xen.hg.
>> Use wmb instead of rmb to enforce ordering between
>> evtchn_upcall_pending and evtchn_pending_sel stores
>> in xen_evtchn_do_upcall().
>>
On Tue, Jun 10, 2008 at 05:41:39PM +1000, Nick Piggin wrote:
> On Tuesday 10 June 2008 17:35, Isaku Yamahata wrote:
> > This patch is ported one from 534:77db69c38249 of linux-2.6.18-xen.hg.
> > Use wmb instead of rmb to enforce ordering between
> > evtchn_upcall_pending and evtchn_pending_sel stor
On native ia64 NET_SKB_NET isn't necessary to define.
However on xen it needs to be slightly larger. (More specifically
it's xen-netfront.c requirement.)
It causes annoying kernel unalgned access message and
network layer troubles.
Resolve it by increasing NET_SKB_NET from the default value 16 to 1
introduce helper functions for xen hypercalls which
traps to hypervisor.
Signed-off-by: Yaozu (Eddie) Dong <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/xen/Makefile |5 +
arch/ia64/xen/hypercall.S| 91 +
include/asm-ia64/x
define several constants for ia64/xen.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/kernel/asm-offsets.c | 27 +++
1 files changed, 27 insertions(+), 0 deletions(-)
diff --git a/arch/ia64/kernel/asm-offsets.c b/arch/ia64/kernel/asm-offsets.c
index c64
Xen needs some address conversions between pseudo physical
address (guest phsyical address), guest machine address
(real machine address) and dma address.
Define helper functions for those address conversion.
Cc: Jeremy Fitzhardinge <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECT
Hi. This patchset implements ia64/xen domU support based on
ia64/pv_ops frame work which was already merged to linux ia64 test branch.
This patch is for linux ia64 test branch.
This patchset does the followings.
- Some preparation work. Mainly importing header files to define
related structures.
eliminate the function declaration ia64_cpu_local_tick() in
process.c by defining in asm-ia64/timex.h
The same function will be used in a different .c file later.
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
arch/ia64/kernel/process.c |1 -
include/asm-ia64/timex.h |2 ++
2 fil
define sync bitops which is necessary for ia64/xen.
Cc: Robin Holt <[EMAIL PROTECTED]>
Cc: Jeremy Fitzhardinge <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
include/asm-ia64/sync_bitops.h | 57
1 files changed, 57 insertions(+
import include/asm-ia64/xen/interface.h to introduce introduce
definitions necessary for ia64/xen hypercalls.
They are basic structures to communicate with xen hypervisor.
Cc: Robin Holt <[EMAIL PROTECTED]>
Cc: Jeremy Fitzhardinge <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED
Xenlinux/ia64 needs to reserve one more region passed from xen hypervisor
as start info.
Cc: Robin Holt <[EMAIL PROTECTED]>
Cc: Bjorn Helgaas <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
include/asm-ia64/meminit.h |3 ++-
1 files changed, 2 insertions(+), 1 deleti
add xen_timer_resume() hook.
Timer resume should be done after event channel is resumed.
add xen_arch_resume() hook when ipi becomes usable after resume.
After resume, some cpu specific resource must be reinitialized
on ia64 that can't be set by another cpu.
However available hooks is run once on o
This patch is ported one from 534:77db69c38249 of linux-2.6.18-xen.hg.
Use wmb instead of rmb to enforce ordering between
evtchn_upcall_pending and evtchn_pending_sel stores
in xen_evtchn_do_upcall().
Cc: Samuel Thibault <[EMAIL PROTECTED]>
Signed-off-by: Isaku Yamahata <[EMAIL PROTECTED]>
---
dr
After restore on ia64 xen domain, kernel panics as follows.
This patch fixes it.
union skb_entry assumes sizeof(link->skb, pointer) ==
sizeof(list->link, unsigned).
However this isn't true on ia64. So make link type unsigned long.
And introduced two accesor.
kernel unaligned access to 0xe000
36 matches
Mail list logo