awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:14:24 +0000 (22:14 -0600)]
[IA64] update buildconfigs for p2m exposure
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:13:06 +0000 (22:13 -0600)]
[IA64] fix sparse tree build with p2m exposure disabled
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:12:50 +0000 (22:12 -0600)]
[IA64] p2m exposure test module
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:12:29 +0000 (22:12 -0600)]
[IA64] p2m exposure. linux side part.
This patch introduce compile time option XEN_IA64_EXPOSE_P2M
to enable this feature and boot option xen_ia64_p2m_expose to
disable the functionality.
This patch also introduce XEN_IA64_EXPOSE_P2M_USE_DTR to map p2m table
with dtr and boot option xen_ia64_p2m_expose_use_dtr to disable it.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:12:25 +0000 (22:12 -0600)]
[IA64] expose p2m table. xen side part.
This patch introduces xen compile time option, xen_ia64_expose_p2m=y.
default is enabled.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:12:14 +0000 (22:12 -0600)]
[IA64] Clean up extern declarations in arch/ia64/xen/domain.c
Signed-off-by: Aron Griffis <aron@hp.com>
awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:12:01 +0000 (22:12 -0600)]
[IA64] remove getmeminfo.nr_pages trick from xencomm
getmeminfo nr_page trick is not anymore necessary.
Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:11:50 +0000 (22:11 -0600)]
[IA64] xencomm_privcmd_sched_op
Necessary for reboot on VTi domains.
Signed-off-by: Kouya Shimura <kouya@jp.fujitsu.com>
awilliam@xenbuild.aw [Thu, 5 Oct 2006 04:11:34 +0000 (22:11 -0600)]
[IA64] revert xen-ia64-unstable.hg cset 11039
No longer necessary with xencomm
Signed-off-by: Kouya Shimura <kouya@jp.fujitsu.com>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 20:46:10 +0000 (14:46 -0600)]
[IA64] add GNTTABOP_copy to xencommize_grant_table_op
This is used w/ xennet.rx_copy
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 14:59:22 +0000 (08:59 -0600)]
merge with xen-unstable.hg
kfraser@localhost.localdomain [Tue, 3 Oct 2006 10:39:22 +0000 (11:39 +0100)]
[NET] Make MAX_SKBUFF_ORDER page-size independent.
Original patch from Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Tue, 3 Oct 2006 10:33:37 +0000 (11:33 +0100)]
[IA64] Fix implicit declaration of initialize_keytable().
Signed-off-by: Aron Griffis <aron@hp.com>
Alastair Tse [Tue, 3 Oct 2006 10:24:48 +0000 (11:24 +0100)]
[XM] Fix rmlabel filename assignment
The patch fixes a bug in the rmlabel tool.
Signed-off-by: Stefan Berger <stefanb@us.ibm.com>
Alastair Tse [Tue, 3 Oct 2006 09:48:07 +0000 (10:48 +0100)]
[XM] Fix missing security check in block-attach
This patch fixes an indentation error in main.py. The effect of this
bug is that block-attach does not check labels if the ACM is
active. This bug slipped in with change set 11572_:_
ad22c711ccb7
<http://xenbits.xensource.com/xen-unstable.hg?cs=
ad22c711ccb7>.
Signed-off by: Reiner Sailer <sailer@us.ibm.com>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 03:53:07 +0000 (21:53 -0600)]
[IA64] backport check_sal_cache_flush()
Backport check_sal_cache_flush() from Linux/ia64. This checks for
SAL_CACHE_FLUSH implementations that drop interrupts and calls
PAL_CACHE_FLUSH instead. This should only be necessary on hp
rx5670 systems.
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 03:42:59 +0000 (21:42 -0600)]
[IA64] don't export GPFN_xxx flags.
GPFN_INV_MASK takes 4 bits, but 1 bit is sufficient.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 03:39:44 +0000 (21:39 -0600)]
[IA64] add perfcounter of dom0vp_phystomach and dom0vp_machtophys
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 03:37:39 +0000 (21:37 -0600)]
[IA64] add perfcounter to mm.c
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 03:34:47 +0000 (21:34 -0600)]
[IA64] add perfcounter for vTLB flush.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 03:32:46 +0000 (21:32 -0600)]
[IA64] avoid long time interrupt masking.
flush_vtlb_fro_context_switch() can be executed with interrupt enabled.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 03:30:51 +0000 (21:30 -0600)]
[IA64] Prevent domains to itc/ptd in shared_info_va.
Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
awilliam@xenbuild.aw [Tue, 3 Oct 2006 03:27:49 +0000 (21:27 -0600)]
[IA64] Bind event channels of VT-i domain to vcpu 0
Signed-off-by: Tsunehisa Doi <Doi.Tsunehisa@jp.fujitsu.com>
awilliam@xenbuild.aw [Mon, 2 Oct 2006 20:09:49 +0000 (14:09 -0600)]
[IA64] Implement fast hypercall for physdevop eoi.
Eoi is a very frequent hypercall which has only one argument passed through
a structure. To avoid the xencomm overhead, a new hypercall is created
and the argument is passed by value.
Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
awilliam@xenbuild.aw [Mon, 2 Oct 2006 20:04:39 +0000 (14:04 -0600)]
[IA64] Renumber dom0vp hypercalls and reformat comments.
(Adding a perfc will be easier with a compact numbering).
Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
awilliam@xenbuild.aw [Mon, 2 Oct 2006 20:03:42 +0000 (14:03 -0600)]
[IA64] Use xencomm for hypercalls.
With xencomm metaphysical addresses are passed to the hypervisor instead
of virtual addresses.
This patch breaks compatibility.
Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
kfraser@localhost.localdomain [Mon, 2 Oct 2006 17:04:56 +0000 (18:04 +0100)]
[HVMLOADER] Clean up SMBIOS table-length computations.
It's easier and less prone to error not to attempt the
length computation up front.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Mon, 2 Oct 2006 15:12:41 +0000 (16:12 +0100)]
[NET] front: Allow copying receive path to be selected by user at
module load time.
Signed-off-by: Keir Fraser <keir@xensource.com>
Steven Smith [Mon, 2 Oct 2006 12:45:44 +0000 (13:45 +0100)]
[XEN][HVM] Fix a bug which could lead to the guest locking up if it
tried to a backwards memcpy across a page boundary in an MMIO region.
This should fix the problems some people have been reporting running
vim.
Signed-off-by: Steven Smith <sos22@cam.ac.uk>
awilliam@xenbuild.aw [Mon, 2 Oct 2006 01:10:18 +0000 (19:10 -0600)]
merge with xen-unstable.hg
awilliam@xenbuild.aw [Sun, 1 Oct 2006 17:39:41 +0000 (11:39 -0600)]
[IA64] warning fix
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
awilliam@xenbuild.aw [Sun, 1 Oct 2006 17:19:45 +0000 (11:19 -0600)]
[IA64] Fix a VTi physical mode bug
When guest writes rr in physical mode, if it is rr0 or rr4, Xen can't
write it into machine rr.
Signed-off-by: Xuefei Xu <anthony.xu@intel.com>
awilliam@xenbuild.aw [Sun, 1 Oct 2006 17:14:00 +0000 (11:14 -0600)]
[IA64] initial xen relocation support
Add support for relocating movl instructions which are run in physical
mode and rely on physical addressing.
This is needed to support platforms which do not provide physical
memory at address 0x4000000.
More relocation points are likely to be needed, please add them as
appropriate in xenpatch.c
Signed-off-by: Jes Sorensen <jes@sgi.com>
awilliam@xenbuild.aw [Sun, 1 Oct 2006 17:09:54 +0000 (11:09 -0600)]
[IA64] do not export XSI_BASE, use set_shared_info_va
Until now, the set_shared_info_va hypercall was never used.
This patch remove XSI_BASE and use this hypercall to set the xsi address.
Because xsi is now in the top 1MB (before percpu space), linux code should be
slightly more compact (addl instead of movl).
Signed-off-by: Tristan Gingold <tristan.gingold@bull.net>
awilliam@xenbuild.aw [Sun, 1 Oct 2006 17:05:24 +0000 (11:05 -0600)]
[IA64] VTi TLB miss fix
When present bit is 0, inject page not present fault to guest.
Signed-off-by: Anthony Xu <anthony.xu@intel.com>
awilliam@xenbuild.aw [Sun, 1 Oct 2006 16:48:40 +0000 (10:48 -0600)]
[IA64] Modify p2m converter to avoid hypervisor crash
Crash occurs during destruction of VT-i domain with PV-on-HVM.
Signed-off-by: Tsunehisa Doi <Doi.Tsunehisa@jp.fujitsu.com>
Signed-off-by: Tomonari Horikoshi <t.horikoshi@jp.fujitsu.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 30 Sep 2006 10:35:02 +0000 (11:35 +0100)]
[HVM][TOOLS] Improve warnign message when trying to create
an HVM guest on a non-capable platform.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 30 Sep 2006 10:30:09 +0000 (11:30 +0100)]
[HVM] Mov to/from CR8 must be gated on whether VLAPIC device is created.
If APIC=0 in VMX configuration file, VLAPIC(v) in
mov_from_cr()/mov_to_cr() will be NULL, so calling
vlapic_get_reg()/vlapic_set_reg()/vlapic_update_ppr()
would crash Xen.
Original patch from Dexuan Cui <dexuan.cui@intel.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 30 Sep 2006 10:11:54 +0000 (11:11 +0100)]
[HVM] Add RDMSR/WRMSR instruction emulation to VMXAssist decoder
AP of PAE SMP windows will use it to set NX bit in EFER.
Signed-off-by: Xin Li <xin.b.li@intel.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 30 Sep 2006 10:08:41 +0000 (11:08 +0100)]
[HVM] Add sub instruction support to HVM MMIO decoder.
This is needed for SMP PAE windows XP SP2.
Signed-off-by: Xin Li <xin.b.li@intel.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 30 Sep 2006 09:49:08 +0000 (10:49 +0100)]
[HVM][SVM] Do not delay ExtInt event injection if RFLAGS.IF==0.
AMD-V does not require the delaying of interrupt injection if the guest
IF_FLAG disallows interrupts. Remove the code in svm_intr_assist()
that was checking the guest IF in eflags.
This patch fixes the problem with HVM Windows guests, with observing a
very slow timer countdown on the initial boot menu, when there is more
than one boot option. Kbd response in this same Windows boot menu is
also acceptably responsive with this patch.
Signed-off-by: Tom Woller <thomas.woller@amd.com>=20
kfraser@localhost.localdomain [Fri, 29 Sep 2006 18:12:15 +0000 (19:12 +0100)]
[NET] back: Loopback must copy foreign mappings.
Signed-off-by: Keir Fraser <keir@xensource.com>
Steven Smith [Fri, 29 Sep 2006 14:40:35 +0000 (15:40 +0100)]
Merge.
Steven Smith [Fri, 29 Sep 2006 14:40:19 +0000 (15:40 +0100)]
[XEN][HVM] Arrange that we only consider sending a PV-on-HVM event channel
callback interrupt in vcpu 0. This avoids a race in SMP HVM domains
which could lead to interrupts never getting delivered. It is safe
because HVM domains always bind their event channels to vcpu 0.
Bug pointed out by Keir.
Signed-off-by: Steven Smith <sos22@cam.ac.uk>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 13:55:57 +0000 (14:55 +0100)]
[BLKTAP] Find major number by searching sysfs
instead of reading /proc/devices.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Steven Smith [Fri, 29 Sep 2006 13:42:49 +0000 (14:42 +0100)]
Merge.
Steven Smith [Fri, 29 Sep 2006 13:42:04 +0000 (14:42 +0100)]
[XEN][HVM] Make sure that the interrupt which event channel events come in
on is level triggered rather than edge triggered, since it's a PCI device.
This is complicated by the possibility that another PCI device could be
on the same interrupt; the workaround is to have two irr registers
for the PIC and APIC, and have qemu and Xen generated interrupts go
into different ones.
This broke the alt_irq stuff. Fortunately, nobody uses that anymore, so
I've removed it.
Signed-off-by: Steven Smith <sos22@cam.ac.uk>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 13:20:52 +0000 (14:20 +0100)]
[LINUX] Make evtchn device use a dynamic minor number.
Also update the code in tools to create the device node if udev fails.
The tools now read the sysfs system to find the minor number needed.
Original patch from Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
Tim Deegan [Fri, 29 Sep 2006 10:57:06 +0000 (11:57 +0100)]
[XEN] Recompute shadows of PAE PSE l2es when needed.
Signed-off-by: Tim Deegan <Tim.Deegan@xensource.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:50:51 +0000 (11:50 +0100)]
[HVM][SVM] Obtaining instruction address needs to mask to 32 bits
if not running in 64-bit mode.
Signed-off-by: Keir Fraser <keir@xensource.com>
Alastair Tse [Fri, 29 Sep 2006 10:25:25 +0000 (11:25 +0100)]
[TOOLS] Fix xendomains init.d script DOG PIG confusion.
Fix typos in xendomains where we actually mean WDOG_PID.
http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=754
Signed-off-by: Alastair Tse <atse@xensource.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:28:04 +0000 (11:28 +0100)]
[HVM] Disable VMX/SVM on auxiliary processors when shutting down an SMP system.
With Intel VT, VMXOFF needs to be executed before halting the
CPU. Without this fix Xen or Linux is unable to start up the CPU after
a kexec reboot.
Tested using x86_32 on a Pentium D 930.
Signed-Off-By: Magnus Damm <magnus@valinux.co.jp>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:26:33 +0000 (11:26 +0100)]
[HVM][VMX] Clear vmxe when VMX is not enabled.
The current Xen code keeps X86_CR4_VMXE set even if VMXON has not been
executed. The stop_vmx() code assumes that it is possible to call
VMXOFF if X86_CR4_VMXE is set which is not always true. Calling VMXOFF
without VMXON results in an illegal opcode trap, and to avoid this condition
this patch makes sure that X86_CR4_VMXE is only set when VMXON has
been called.
Tested using x86_32 on a Pentium D 930.
Signed-Off-By: Magnus Damm <magnus@valinux.co.jp>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:22:24 +0000 (11:22 +0100)]
[PYGRUB] Enable showing the cursor in pygrub when it's relevant to do so.
Signed-off-by: Jeremy Katz <katzj@redhat.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:21:02 +0000 (11:21 +0100)]
[HVM][SVM] 64-bit Windows HVM guests require MCE/MCA CPUID bits to be present.
This SVM patch removes "masking off" of these bits for AMD-V HVM guests.
Signed-off-by: Tom Woller <thomas.woller@amd.com>=20
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:17:29 +0000 (11:17 +0100)]
[BLKTAP] Remove unnecessary TLB flush from blktap driver.
blktap_poll is calling tlb_flush_all() in its main ring buffer polling
loop. This seems to be superfluous: the hypervisor should be
performing
any necessary tlb flushes on grant table operations performed by the
back-end. Even a simple memory barrier is unnecessary here as the
RING_PUSH_REQUESTS() call performs a wmb() anyway.
And tlb_flush_all() is not exported to modules, so this call prevents
blktap from building as a module. Just remove it.
Signed-off-by: Stephen Tweedie <sct@redhat.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:16:52 +0000 (11:16 +0100)]
[BLKTAP] Remove unnecessary timeout from tapdisk select() call.
tapdisk was using a 1msec select() timeout in its main event loop. On
timeout, the loop does nothing except repeat, so the timeout achieves
nothing; worse, on a machine with HZ set to less than 1000, a msec
timeout returns immediately, sending tapdisk into a 100%-CPU-
utilisation loop.
Simply remove the timeout entirely; everything still works fine, and
much more efficiently when idle.
Signed-off-by: Stephen Tweedie <sct@redhat.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:13:01 +0000 (11:13 +0100)]
Remove dead pdb code from tools directory.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:11:49 +0000 (11:11 +0100)]
[VMXASSIST] Fix data32/addr32 instruction decode issue.
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 10:10:14 +0000 (11:10 +0100)]
[HVM] Clean up hvm_copy interface to be more like copy_to/from_user.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 09:30:18 +0000 (10:30 +0100)]
[HVM][VMX] Fix data copying in transition to/from vmxassist.
In vmx_assist, the copy for new/old context and vmx assist magic are
all using physical address, while hvm_copy will use the virtual address.
This may cause problem when guest jump directly from real mode to
paging mode.
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 08:29:20 +0000 (09:29 +0100)]
[NET] back: Reset netfilter context in loopback receive path.
The issue is described in bugzilla report #746.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 08:23:18 +0000 (09:23 +0100)]
[BLK] front: Flush workqueues with no locks held. Operation can sleep.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Fri, 29 Sep 2006 08:14:28 +0000 (09:14 +0100)]
[XENBUS] Don't perform normal state-machine transitions and
notifications during shutdown. When rootfs goes away this can
cause the system to hang, as we notify userspace etc.
Instead we forcibly reset the state machine back to state Closed
if we are a frontend driver, to get the backend to release
resources. Necessary if we are going to kexec.
Signed-off-by: Keir Fraser <keir@xensource.com>
Andrew Warfield [Thu, 28 Sep 2006 20:44:00 +0000 (13:44 -0700)]
[BLKTAP] have udev create the device for blktap
This patch makes blktap Do The Right Thing(TM). It allows udev to
create the /dev/xen/blktap[0-9] devices.
It creates a sysfs class called "xen". This part may later be placed
someplace else, but currently blktap is the only user so it is placed in
the blktap code.
When blktap is initialized, a blktap0 sysfs class device is made. The
other devices blktapX (X > 0) are made when the BLKTAP_IOCTL_NEWINTF
ioctl is called. This way we don't flood the sysfs and /dev/xen with
unnecessary devices.
I added a rule in the xen-backend.rules to allow for udev to create the
blktap devices.
With this, we can really remove the code to search and create the
/dev/xen/blktap[0-9]*, but I'll leave it in for now. With the use of
udev, we really should remove that code as well as the code for creating
the evtchn device. udev works for both of these now. But that removal
will have to be in another patch.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Andrew Warfield [Thu, 28 Sep 2006 19:47:45 +0000 (12:47 -0700)]
[BLKTAP] have blktap use a dynamic major
blktap currently uses a hardcoded major of 254 for the device. This is
not robust in anyway and needs to be dynamic.
Note: it is better not to have the daemon create the node, and have udev
create it instead. But since the daemon currently creates the node
anyway, it is still the way this is done. That change needs to be made
at another time.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Andrew Warfield [Thu, 28 Sep 2006 19:12:59 +0000 (12:12 -0700)]
[BLKTAP][XEND] Fix blktap to work with a bootloader
If a Xen guest has a bootloader configured, then it will fail to start
on a blktap image. The problem is blkdev_uname_to_file, which cannot
parse the "tap:aio:$filename" image strings: it tries to split the
string apart at ":" and assign the result to a 2-tuple, and this
results in a python error if the split results in three or more
strings.
The fix is to split only at the first ":", and then to split again
if we detect "tap:" as the image type.
Signed-off-by: Stephen Tweedie <sct@redhat.com>
Andrew Warfield [Thu, 28 Sep 2006 19:03:01 +0000 (12:03 -0700)]
[BLKTAP] Fix SMP debugging assert failures in blktap
blktap is calling non-atomic kernel functions with irqs disabled, which
hits various kernel debug traps if those are enabled. The problem is
req_increase(), which takes the pending_free_lock.
This function is currently only ever called from module initialiation,
where it is impossible for it to race against anything else. Its
companion, req_decrease(), is not called at all.
Fix by removing the offending locking from req_increase() and, while
we're at it, remove req_decrease() entirely.
Signed-off-by: Stephen Tweedie <sct@redhat.com>
Andrew Warfield [Thu, 28 Sep 2006 18:52:17 +0000 (11:52 -0700)]
[BLKTAP] clean up blktap and remove private structure
This patch cleans up the blktap.c code to make it form to the Linux
coding style a little better.
It also removes the private data structure that is only used to store
the index of the tabfds descriptor. Instead the filp->private_data now
points to the descriptor itself.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Andrew Warfield [Thu, 28 Sep 2006 18:41:23 +0000 (11:41 -0700)]
[BLKTAP] Fix blktap oops on domain shutdown
When a domain shuts down with a blktap-backed block device open, it can
easily cause a dom0 oops. The XenbusStateClosing event can occur while
the tapdisk userland thread is still processing IO requests (eg.
readaheads) from the domU. But the xenbus state handler calls
tap_blkif_unmap(), unmapping the blkif->blk_ring.sring rin buffer, so
when the tapdisk thread next calls the BLKTAP_IOCTL_KICK_FE to return
the completion event to the FE via that ring buffer, it oopses.
This can be fixed simply by not calling tap_blkif_unmap() in this case;
the ring buffer will still be unmapped later on when the blkif is
destroyed by blktap_remove(), only then it will properly wait for the
blkif refcnt to reach zero before doing so.
Signed-off-by: Stephen Tweedie <sct@redhat.com>
Tim Deegan [Thu, 28 Sep 2006 16:10:54 +0000 (17:10 +0100)]
[XEN] Support lightweight shadow-translate PV guests, for paravirt-ops.
This is a modified subset of Michael Fetterman's shadow-translate work.
Signed-off-by: Tim Deegan <Tim.Deegan@xensource.com>
Tim Deegan [Thu, 28 Sep 2006 16:09:11 +0000 (17:09 +0100)]
[XEN] Fix interaction between tlbflush timestamp and shadow flags
Signed-off-by: Tim Deegan <Tim.Deegan@xensource.com>
kfraser@localhost.localdomain [Thu, 28 Sep 2006 12:45:49 +0000 (13:45 +0100)]
[NET] back: 4 loopbacks by default for dom0 (previously was 8).
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Thu, 28 Sep 2006 12:44:24 +0000 (13:44 +0100)]
[XEN] Remove weak-extern definitions from symbols.c.
Instead we have an explicit set of dummy symbol
definitions for the inital link of the Xen image.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Thu, 28 Sep 2006 12:20:58 +0000 (13:20 +0100)]
[NET] back: Create 0 loopbacks by default for non-domain-0.
Signed-off-by: Keir Fraser <keir@xensource.com>
atse@norwich.uk.xensource.com [Thu, 28 Sep 2006 11:46:09 +0000 (12:46 +0100)]
[XM] Report proper ACMError(s) instead of silently exiting.
Remove try except blocks so that ACMError(s) are properly reported to
the user rather than silently fail.
Signed-off-by: Alastair Tse <atse@xensource.com>
atse@norwich.uk.xensource.com [Thu, 28 Sep 2006 11:29:52 +0000 (12:29 +0100)]
[XM] Remove traceback printing from load/makepolicy.py.
Should pass the exception along to xm/main.py to handle properly like
the rest of the commands.
Signed-off-by: Alastair Tse <atse@xensource.com>
atse@norwich.uk.xensource.com [Thu, 28 Sep 2006 11:23:21 +0000 (12:23 +0100)]
[XM] Fix stray quotes in usage message in getlabel.py
Fixed built-in function name conflict.
Raise correct exceptions for when option is invalid to be properly
reported by xm.
Signed-off-by: Alastair Tse <atse@xensource.com>
atse@norwich.uk.xensource.com [Thu, 28 Sep 2006 11:16:11 +0000 (12:16 +0100)]
[XM] Fix error reporting for resources.py, remove duplicate try block.
Also fixed usage message to not include misspelling of the command.
Signed-off-by: Alastair Tse <atse@xensource.com>
atse@norwich.uk.xensource.com [Thu, 28 Sep 2006 10:50:51 +0000 (11:50 +0100)]
[XM] Removed traceback printing for non ACMError exceptions
Removing try,except block as cfgbootpolicy should allow exceptions to
propagate instead of printing traceback to stdout.
(http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=778)
Changed usage of string.join, string.split when creating security
policy path name.
Signed-off-by: Alastair Tse <atse@xensource.com>
atse@norwich.uk.xensource.com [Thu, 28 Sep 2006 11:03:10 +0000 (12:03 +0100)]
[XM] Set proper return code for exceptions in xm, catch ACMError.
Signed-off-by: Alastair Tse <atse@xensource.com>
kfraser@localhost.localdomain [Thu, 28 Sep 2006 10:47:33 +0000 (11:47 +0100)]
[XEN] Fix i386 hypercall code to always create shadow
parameters. Arguments on stack can be clobbered by
callee, since it owns them. But we require this doesn't
happen when we create hypercall continuations. Hence the
need to copy.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Thu, 28 Sep 2006 10:25:15 +0000 (11:25 +0100)]
[HVM] Avoid mis-optimisation of the SVM/VMX exit handlers
by passing a *pointer* to the register state, instead of
'passing by value'.
This should fix the bugs resulting in display of
"trying to set reserved bit in EFER"
on the Xen console.
Signed-off-by: Keir Fraser <keir@xensource.com>
atse@norwich.uk.xensource.com [Thu, 28 Sep 2006 09:10:59 +0000 (10:10 +0100)]
There's a typo in create.py, which affects PV bootloaders (such as
domUloader and probably pygrub). Patch attached.
Signed-off-by: Charles Coffing <ccoffing@novell.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 16:49:22 +0000 (17:49 +0100)]
[VMXASSIST] Cast to pointer type via long to avoid compile
warnings/errors with latest gcc versions.
Signed-off-by: Keir Fraser <keir@xensource.com>
Steven Smith [Wed, 27 Sep 2006 15:22:17 +0000 (16:22 +0100)]
[HVM] qemu should wait 10ms, rather than 100ms, between running its timers.
Signed-off-by: Steven Smith <sos22@cam.ac.uk>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 13:30:36 +0000 (14:30 +0100)]
[XM-TEST] Remove the test for /dev/tpm0, which may be a prerequisite to start
the vtpm_manager. A 'SKIP' is shown if the vtpm_manager process has not been started.
Signed-off-by: Stefan Berger <stefanb@us.ibm.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 13:28:26 +0000 (14:28 +0100)]
[XEN] gdbstub return value is used to determine whether or not
to continue execution.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 13:01:30 +0000 (14:01 +0100)]
[XM] Fix 'xm uptime' command.
Signed-off-by: Masaki Kanno <kanno.masaki@jp.fujitsu.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 13:00:33 +0000 (14:00 +0100)]
[XM] Fix domain reported by 'xm sched-sedf'.
Signed-off-by: Masaki Kanno <kanno.masaki@jp.fujitsu.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 12:59:48 +0000 (13:59 +0100)]
[XM] Fix 'xm rmlabel' command.
Signed-off-by: Masaki Kanno <kanno.masaki@jp.fujitsu.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 12:58:46 +0000 (13:58 +0100)]
[XM] Reenable the '-h' option, disabled in changeset:11572.
Signed-off-by: Masaki Kanno <kanno.masaki@jp.fujitsu.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 12:54:04 +0000 (13:54 +0100)]
[HVM] Fix stupid printf formatting error that breaks x86/64 build.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 09:17:03 +0000 (10:17 +0100)]
[HVM] Fix MSR read/write and add extra info when
write to MSR_EFER fails.
Signed-off-by: Keir Fraser <keir@xensource.com>
kfraser@localhost.localdomain [Wed, 27 Sep 2006 08:29:46 +0000 (09:29 +0100)]
[XEN] Fix x86_emulate and hvm-mmio-insn-len decoders
to properly add cs<<4 to eip only when guest is in real
mode (or vm86 mode).
Remove bogus test-and-fail from hvm-mmio-insn-len decoder.
Signed-off-by: Keir Fraser <keir@xensource.com>
awilliam@xenbuild.aw [Wed, 27 Sep 2006 01:11:33 +0000 (19:11 -0600)]
merge with xen-unstable.hg
awilliam@xenbuild.aw [Tue, 26 Sep 2006 22:15:45 +0000 (16:15 -0600)]
[IA64] Complete fpswa handler retry mechanism
When handling fpswa fault, Xen needs to fetch opcode, it may fail.
This patch finishes retry mechanism.
Signed-off-by: Anthony Xu <anthony.xu@intel.com>
Steven Smith [Tue, 26 Sep 2006 18:50:07 +0000 (19:50 +0100)]
[HVM][VNC] Fix typo.
Signed-off-by: Steven Smith <sos22@cam.ac.uk>
Steven Smith [Tue, 26 Sep 2006 15:47:35 +0000 (16:47 +0100)]
[HVM][VNC] Work around a bug in the newest version of xvncviewer which
could lead to the display locking up after it's been idle for a while.
Signed-off-by: Steven Smith <sos22@cam.ac.uk>
Steven Smith [Tue, 26 Sep 2006 15:46:47 +0000 (16:46 +0100)]
[HVM][VNC] Make sure that qemu doesn't go into an infinite loop when
it receives certain invalid requests from the viewer.
Signed-off-by: Steven Smith <sos22@cam.ac.uk>