kaf24@firebug.cl.cam.ac.uk [Sat, 1 Apr 2006 09:49:42 +0000 (10:49 +0100)]
Fix makefiles for hvmloader and vmxassist for FC5 on x86_64
systems.
In FC5, gcc has issues with the command lines used to build
hvmloader and vmxassist. Basically, the compiler doesn't
like the combination of -m64 and -m32. As these are 32 bit
apps, and intended to be, take a modified approach and simply
define XEN_TARGET_ARCH for 32 bits. This cleans up the
command line and builds the images correctly.
From: Ben Thomas <ben@virtualiron.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 1 Apr 2006 09:40:54 +0000 (10:40 +0100)]
[SVM] Do not update seg.base in realmode while obtaining the io
addr for ins/outs.
Signed-off-by: Tom Woller <thomas.woller@amd.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 1 Apr 2006 09:39:31 +0000 (10:39 +0100)]
Minor minios traps.c fix for x86/64.
Signed-off-by: Aravindh Puthiyaparambil
<aravindh.puthiyaparambil@unisys.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 1 Apr 2006 09:37:36 +0000 (10:37 +0100)]
[IA64] Include features header in xenbus_probe and build features.c
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 1 Apr 2006 09:20:10 +0000 (10:20 +0100)]
Add new boot parameter 'lowmem_emergency_pool' which creates
an emergency pool of pages below 4GB (i.e., in Xen's DMA pool)
which will ever be used to satisfy normal allocation requests.
This is particularly useful in an i386 PAE environment to ensure
that pages will always be available for allocation as
page-table base directories.
Usage, for example (to reserve 16MB):
lowmem_emergency_pool=16M
In a PAE environment, each reserved megabyte guarantees you should
be able to create up to 256 processes (counted across all guests).
16MB means you're good for 4096 processes.
Signed-off-by: Keir Fraser <keir@xensource.com>
cl349@firebug.cl.cam.ac.uk [Fri, 31 Mar 2006 16:44:26 +0000 (17:44 +0100)]
Fix do_IRQ high bit masking.
Instead of setting the highest bit (which isn't easily done on native x86_64),
negate the interrupt vector stored in orig_{e,r}ax.
Also add patch for native build.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
kaf24@firebug.cl.cam.ac.uk [Fri, 31 Mar 2006 14:34:52 +0000 (15:34 +0100)]
Plumb network vif credit-based rate limiting thorugh xenbus
and xend into xm guest config files.
A new vif parameter 'rate' is supported, with an optional time window
paremeter for specifying granularity of credit replenishment. The default
window is 50ms. For example:
'rate=10Mb/s' 'rate=250KB/s' 'rate=1MB/s@20ms'
From: Chris Clark <christopher.w.clark@gmail.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 31 Mar 2006 12:51:19 +0000 (13:51 +0100)]
Currently, xm & xend allow an XAUTHORITY to be specified in the config
file. If that is not specified, then xm will pass the XAUTHORITY
environment variable back to xend instead. This is not sufficient,
since it is possible that XAUTHORITY is not set at all. This doesn't
cause a problem for most X apps, because xlib will use the fall-back of
$HOME/.Xauthority. But the viewer is started from xend, which almost
certainly doesn't have a reasonable HOME variable.
We fix this by always passing some value for
XAUTHORITY. In order of preference, that value is:
1. Use the value from the config file, if specified.
2. Use the value from the XAUTHORITY env variable, if set.
3. Use $HOME/.Xauthority, if HOME is set.
4. Look up the user's home directory via getpwuid, and append
"/.Xauthority"
Signed-off-by: Charles Coffing <ccoffing@novell.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 31 Mar 2006 11:03:42 +0000 (12:03 +0100)]
Various softirq cleanups:
1. Make __softirq_pending a long, since PPC borrows the file
hardirq.h from us and they only do atomic ops on longs.
2. do_softirq() explicitly takes a void param list.
3. Remove idle_timestamp field from irq_cpustat. It's unused
and lets us simplify the idle loop a little bit.
Part 1 based on a patch from Hollis Blanchard at IBM.
Signed-off-by: Keir Fraser <keir@xensource.com>
Ian.Campbell@xensource.com [Fri, 31 Mar 2006 10:04:42 +0000 (11:04 +0100)]
Disable xen bus and grant tables when supervisor_mode_kernel is enabled.
Neither of these features are useful/available in this mode since only a
single domain is supported.
Do not attempt to initialise xen bus when supervisor_mode_kernel is
enabled.
Do not BUG_ON() failure to setup grant tables, future versions
of supervisor_mode_kernel may return -ENOSYS here.
Signed-off-by: Ian Campbell <ian.campbell@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 31 Mar 2006 09:24:17 +0000 (10:24 +0100)]
See panic messages before the post message processing.
Signed-off-by: Jimi Xenidis <jimix@watson.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 31 Mar 2006 09:07:55 +0000 (10:07 +0100)]
Enable the setting and trapping of breakpoints for hvm guest.
Catch Ctrl-C for gdbserver and let gdb break from continue command.
Signed-Off-By: Nitin A Kamble <nitin.a.kamble@intel.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 31 Mar 2006 09:06:20 +0000 (10:06 +0100)]
The ia64 build doesn't yet include drivers/xen/Kconfig directly so we
need to replicate the new XEN_INTERFACE_VERSION option in the arch
Kconfig.
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 23:26:07 +0000 (00:26 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 23:24:54 +0000 (00:24 +0100)]
Remove the transaction parameter from xenbus_switch_state and move the state
switch out of a transaction, in the few cases where it is inside one.
In order to behave properly, it is necessary for a driver to know its own
xenbus state (see changeset 9469:
b3cb19d2b07f, for example). This
value is stored as xenbus_device.state and updated by xenbus_switch_state.
If xenbus_switch_state occurs within a transaction, then there is a possibility
that the transaction would be aborted, leaving the state field dangerously out
of sync with the value currently in the store.
This fixes recent problems seen whereby bringing multiple devices up at the
same time results in some devices not coming up (often all of the even-numbered
ones, because of the pattern of transaction conflict).
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 23:15:12 +0000 (00:15 +0100)]
Better formatting of error messages for xmlrpclib.Faults, matching similar code
in main.py.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 23:14:09 +0000 (00:14 +0100)]
Print error messages to stderr. Remove unused handle_xend_error.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 23:13:33 +0000 (00:13 +0100)]
Set the permissions correctly on the XML-RPC UDP socket, so that non-root users
cannot use the socket.
This closes a security hole, and fixes the intermittent failure
of xm-test/06_list_nonroot.test.
c.f. xen-unstable changeset 9205:
faa1eb1621b9 (same bug, different socket).
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 23:10:54 +0000 (00:10 +0100)]
Further attempts to recover from a corrupt store, this one triggered when a
tdb_store fails.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 23:09:42 +0000 (00:09 +0100)]
Don't use abbreviated ip subcommands -- these are not accepted by iproute2.
Closes bug #478.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
cl349@firebug.cl.cam.ac.uk [Thu, 30 Mar 2006 16:50:44 +0000 (17:50 +0100)]
Remove __HYPERVISOR_sched_op_new in favour of a header file interface version number system.
Define interface version in Kconfig and define it through CPPFLAGS for
C and assembly file compilation.
Add HYPERVISOR_{yield,block,shutdown} to remove direct use of sched_op
from guest code.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
cl349@firebug.cl.cam.ac.uk [Thu, 30 Mar 2006 16:50:40 +0000 (17:50 +0100)]
Change how we add the include asm override to CPPFLAGS.
Add to CPPFLAGS instead of entirely redefining it.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 16:26:19 +0000 (17:26 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 16:25:03 +0000 (17:25 +0100)]
Added medium-length (around 20 minute run-time) test group.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 30 Mar 2006 15:53:37 +0000 (16:53 +0100)]
The attached trivial patch fixes cases where a block device is mounted
read-only in domain 0, but still fails to be shared with a domU, because
it is mis-identified as being mounted writable. (On SUSE, CDs and DVDs
are auto-mounted not with merely permissions "ro", but
"ro,nosuid,nodev", hence the mis-identification.)
Signed-off-by: Charles Coffing <ccoffing@novell.com>
smh22@firebug.cl.cam.ac.uk [Thu, 30 Mar 2006 14:28:53 +0000 (15:28 +0100)]
Fix comment.
Signed-off-by: Steven Hand <steven@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 30 Mar 2006 13:37:22 +0000 (14:37 +0100)]
More fixes to gdbserver for HVM guest debugging. Also fix
writing back of register state for HVM guests.
From: Nitin A Kamble <nitin.a.kamble@intel.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 30 Mar 2006 13:31:57 +0000 (14:31 +0100)]
Use AFLAGS for assembly files so we can "+=" more flags.
From: Jimi Xenidis <jimix@watson.ibm.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 30 Mar 2006 13:06:11 +0000 (14:06 +0100)]
Include the E7520 (e.g., Dell 1850) irqbalance quirk fix even when
building a non-SMP kernel. The underlying hypervisor may still be SMP.
Signed-off-by: Keir Fraser <keir@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 10:55:07 +0000 (11:55 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 10:54:07 +0000 (11:54 +0100)]
Check the return value of domain_lookup_by_name_or_id_nr for None (i.e. no
such domain) inside each of the public-facing functions. This fixes the
Internal errors seen when specifying an invalid domain.
domain_pincpu takes cpumap, which is a list, not a string, so there's no need
to try and split it up. Fixes xm vcpu-pin.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 30 Mar 2006 10:51:44 +0000 (11:51 +0100)]
Fix the ProtocolError seen when the server throws an exception and running
under Python 2.3; traceback.format_exc was introduced in 2.4, so we can't use
it.
Added some exception logging.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 30 Mar 2006 10:38:56 +0000 (11:38 +0100)]
Fix xenconsoled when sending lots of console data to a domU.
Fixes bug #477.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 23:11:53 +0000 (00:11 +0100)]
Same panic() behaviour as native Linux when running as domain 0.
That is, no reboot: just sit there and flash the keyboard LEDs.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 23:07:23 +0000 (00:07 +0100)]
Xen waits 5 seconds to reboot when domain 0 crashes, giving
time to read crash messages.
From: Chris Wright <chrisw@sous-sol.org>
Signed-off-by: Keir Fraser <keir@xensource.com>
emellor@leeni.uk.xensource.com [Wed, 29 Mar 2006 17:46:51 +0000 (18:46 +0100)]
Revert accidental commit.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Wed, 29 Mar 2006 17:41:30 +0000 (18:41 +0100)]
Merged.
smh22@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 17:07:36 +0000 (18:07 +0100)]
Fix save/restore on 64-bit.
Signed-off-by: Steven Hand <steven@xensource.com>
smh22@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 17:06:27 +0000 (18:06 +0100)]
Fix save/restore bug; further rationalization of xenbus state machine
logic deferred until post 3.0.2.
Signed-off-by: Steven Hand <steven@xensource.com>
cl349@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 15:50:59 +0000 (16:50 +0100)]
Free shadows of any pages which are released by a domain back to Xen.
This should fix the "Assertion '! IS_INVALID_M2P_ENTRY(gpfn)'"-crash in
free_shadow_page on domain destruction.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
cl349@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 15:47:46 +0000 (16:47 +0100)]
Re-arrange code for followup patch and remove extra shadow_lock in function which is only called with the lock already held.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
kaf24@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 15:02:40 +0000 (16:02 +0100)]
Compute actual baud rate from UART divisor latch contents
when no baud rate is specified. Generalise the divisor
calculation based on external clock rate.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 14:39:22 +0000 (15:39 +0100)]
Ensure curr_vcpu in domain.c is set correctly, even when
nr physical cpus is greater than max virtual cpus per domain.
Also do not initialise secondary CPU smp_processor_id() from
smpboot.c cpucount. It will be wrong if some CPUs fail to boot.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 13:54:43 +0000 (14:54 +0100)]
Make maximum number of supported physical CPUs a compile-time
option via the 'max_phys_cpus=<nr>' compilation parameter.
Based on a patch from Aravindh Puthiyaparambil at Unisys.
Signed-off-by: Keir Fraser <keir@xensource.com>
anthony@rhesis.austin.ibm.com [Wed, 29 Mar 2006 13:50:57 +0000 (14:50 +0100)]
Remove vcpu_avail from the public S-Expression that's passed over the wire.
This trips up the XML-RPC layer on large SMP systems and isn't actually used
by xm.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 12:56:26 +0000 (13:56 +0100)]
SVM patch to update guest time with latest hvm modifications.
Signed-off-by: Tom Woller <thomas.woller@amd.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 11:02:45 +0000 (12:02 +0100)]
Some fixes to IRET hypercall and failsafe callback handlers:
1. IRET hypercall must restore the event callback mask. This
was missing on x86/64, and both subarchitectures now restore
from EFLAGS.IF.
2. Failsafe callbacks are fixed to detect whether fault is due to
bad segment or due to IRET. In the latter case we now
immediatiately kill the process. This avoids infinite looping
between IRET hypercall and failsafe callback handler in the guest.
Signed-off-by: Keir Fraser <keir@xensource.com>
cl349@firebug.cl.cam.ac.uk [Wed, 29 Mar 2006 09:46:36 +0000 (10:46 +0100)]
Initialise p2m entries to INVALID_MFN and disable debug printk in pfn to mfn lookup.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
kaf24@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 20:21:26 +0000 (21:21 +0100)]
Fix get_page_type() when passed PGT_va_mutable.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 17:43:30 +0000 (18:43 +0100)]
Clean up and fix VCPU hotplug and SMP save/restore.
1. No longer hold xenbus_lock while taking down VCPUs
in SMP suspend path. This allows block device hotplug
to continue working and so we will not deadlock on
paging in userspace hotplug code.
2. Track xenbus and local-admin permitted cpumasks for
VCPUs to bring online. So, if a local admin takes a
CPU down, that won't surprisingly get overridden next
time the kernel interrogates xenstore.
Signed-off-by: Keir Fraser <keir@xensource.com>
cl349@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 13:19:22 +0000 (14:19 +0100)]
Change do_IRQ high bit masking.
Allow more than 256 interrupt vectors on native by only setting the
highest bit when marking orig_eax to indicate that we're not within
a system call.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
kaf24@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 12:45:08 +0000 (13:45 +0100)]
The introduction of the SMI handling code added in cset 9371 is
"reversed", as in it actually clears the SMI bit in the VMCB. If you
have a new enough chip (or new BIOS), the SMI bit HAS TO BE SET or it
gives exitcode -1.
Signed-off-by: Mats Petersson <mats.petersson@amd.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 12:43:22 +0000 (13:43 +0100)]
merge
kaf24@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 12:42:28 +0000 (13:42 +0100)]
Add missing -xen ia64 config file.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 12:40:30 +0000 (13:40 +0100)]
SVM only patch to fix location of PIO request RIP update.
Fixes booting of win2003 server, and various CD installation failures.
Signed-off-by: Tom Woller <thomas.woller@amd.com>
emellor@leeni.uk.xensource.com [Tue, 28 Mar 2006 10:52:47 +0000 (11:52 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Tue, 28 Mar 2006 10:51:06 +0000 (11:51 +0100)]
Mention uClibc and buildroot configs.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Tue, 28 Mar 2006 10:47:58 +0000 (11:47 +0100)]
Added README for the XenSource-hosted initrd.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Tue, 28 Mar 2006 10:47:20 +0000 (11:47 +0100)]
Download the initrd from xm-test.xensource.com, unless otherwise specified.
Don't download it again if it's already present.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Tue, 28 Mar 2006 10:26:29 +0000 (11:26 +0100)]
Watch for @introduceDomain, for symmetry with the existing watch on
@releaseDomain.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Tue, 28 Mar 2006 10:25:22 +0000 (11:25 +0100)]
Have TCPXMLRPCServer inherit from SocketServer.ThreadingMixIn. This allows
Xend to handle more than one request simultaneously once again.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
cl349@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 09:09:44 +0000 (10:09 +0100)]
Remove special handling of {set,clear}_in_cr4().
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
kaf24@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 09:01:35 +0000 (10:01 +0100)]
Simplify the failsafe callback handler in x86/64 linux. It doesn't
need to try reloading segment selectors (unlike i386 linux).
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 28 Mar 2006 08:57:48 +0000 (09:57 +0100)]
Add -xen buildconfig for ia64 and tweak CONFIG_VT setup to avoid
initializing on domUs. Remove CONFIG_IDE_GENERIC as this is
unnecessary on ia64 systems (no ISA IDE controllers) and causes long
timeouts booting domU.
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
kaf24@firebug.cl.cam.ac.uk [Mon, 27 Mar 2006 17:03:03 +0000 (18:03 +0100)]
Make -xen default kernel compile on x86_64 by removing b44 module.
This patch ends up affecting a couple other lines because it is the
result of a new make menuconfig, but those changes should be harmless.
Signed-off-by: Sean Dague <japh@us.ibm.com>
cl349@firebug.cl.cam.ac.uk [Mon, 27 Mar 2006 10:16:36 +0000 (11:16 +0100)]
Whitespace.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
kaf24@firebug.cl.cam.ac.uk [Sun, 26 Mar 2006 10:50:39 +0000 (11:50 +0100)]
Fix 64-bit HVM guest debug via gdbserver:
1. gdb_regs for 64bit, and macros to transfer regisers between ptrace
registers to gdb registers
2. xc_ptrace code building for 64bit
3. Implementation of new map_domain_va for 64bit
4. gdbserver-xen build configuration fixes for 64bit
From: Nitin Kamble
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Sun, 26 Mar 2006 10:45:35 +0000 (11:45 +0100)]
Allow 64-bit Xen to run 64-bit hvm SMP guests.
Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
kaf24@firebug.cl.cam.ac.uk [Sun, 26 Mar 2006 10:43:53 +0000 (11:43 +0100)]
Don't reclaim vcpus that are not put into use yet when destroying HVM
domain. Otherwise, when running a debug=y xen, we will get a assertion
failure in vmx_request_clear_vmcs(), since some vcpus may be not
initialized for HVM use yet.
Signed-off-by: Xin Li <xin.b.li@intel.com>
Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
kaf24@firebug.cl.cam.ac.uk [Sun, 26 Mar 2006 10:42:17 +0000 (11:42 +0100)]
Fix build error with debug=y inxen/ia64 by moving
MAX_ORDER definition has moved into a header file.
Signed-off-by: Masaki Kanno <kanno.masaki@jp.fujitsu.com>
kaf24@firebug.cl.cam.ac.uk [Sun, 26 Mar 2006 10:39:04 +0000 (11:39 +0100)]
The attached patch to the qemu emulation of the pcnet hardware fixes
several problems. It will now only read and write a transmit or receive
descriptor once. It will correctly handle transmitting frames with more
than two fragments. It will discard oversize frames instead of
corrupting memory. I have tested all the changes I have made and even
seen an improvement in receive performance from 1.7MB/s to 3.3MB/s.
Your mileage will vary.
The code could be simplified if multi-fragment frames were deleted. It
appears that both Linux and Windows XP don't use fragmented frames.
The mac crc computation code was deleted, since it has never been used.
The code was checking the wrong bit in the control register.
I tested type 3 descriptors by modifying the linux pcnet32 driver, but
did not see any difference in performance over type 2 currently used.
I have not made any change to type 0 (16-bit) transmit/receive
descriptors as I have no way to test the changes.
Please test this as soon as possible, as I will be on vacation next week
and in a class the following week. I will check email from time to
time, but my ability to make changes to code will be greatly dimished.
One other thing I noticed in my testing, is that if I tell a linux domU
to reboot, it will reboot but not have any pcnet32 device after
rebooting. By doing a shutdown, xm destroy and xm create, I was able to
get around that problem.
Signed-off-by: Don Fry <brazilnut@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Sun, 26 Mar 2006 10:35:11 +0000 (11:35 +0100)]
Add format printf attribute to panic() prototype and fix the error
this catches.
Signed-off-by: Jimi Xenidis <jimix@watson.ibm.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 24 Mar 2006 16:48:03 +0000 (17:48 +0100)]
emellor@leeni.uk.xensource.com [Fri, 24 Mar 2006 13:36:14 +0000 (14:36 +0100)]
Switch the default build to make the -xen kernel, not the -xen0 and -xenU
kernels. Distros will be using -xen exclusively, so this change means that
developers and users of unstable will be (by default) using the same
configuration as the distros.
This can be overridden by setting KERNELS="linux-2.6-xen0 linux-2.6-xenU" on the
command line.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
harry@localhost.localdomain [Fri, 24 Mar 2006 13:31:20 +0000 (14:31 +0100)]
Fix bug 515 by adding a global lock around the hotplug scripts in the non-udev hotplug case only.
There are two possible root causes for 515, both of which are only problems in the non-udev hotplug case because udev uses udevsend which already implements the required serialisation.
1) Script concurrency.
2) Kernel reordering hotplug events.
This changeset fixes (1) but not (2). Since (1) is the problem that seems to be happening this is probably OK.
A fix for (2) for the hotplug case might be to add extra serialisation using state changes in the store but this would impact all the drivers and the code and extra complexity would be redundant once everyone had moved to udev. This doesn't seem worthwhile unless we actually start to see (2) happening.
anthony@rhesis.austin.ibm.com [Fri, 24 Mar 2006 13:31:12 +0000 (14:31 +0100)]
1) Introduce new exception type XendInvalidDomain that maps to the high level
XEND_INVALID_DOMAIN faultType.
2) Fix exception logic in XMLRPCServer
3) Fix TCP server
4) Remove catching of ProtocolError in main.py. ProtocolErrors only occur
when there is an exception in the exception handling code which shouldn't
ever happen. I've reproduced the error cases described by Ewan with
xend_domain_setTargetMemory and once I fixed the exception logic, I get a
normal faultType of 1 as would be expected.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 24 Mar 2006 11:28:58 +0000 (12:28 +0100)]
Create /var/xen/dump/ during tools installation, or core files
are not written even if we set (enable-dump yes) in xend-config.sxp.
Signed-off-by: Akio Takebe <takebe_akio@jp.fujitsu.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 24 Mar 2006 11:14:58 +0000 (12:14 +0100)]
Detect spurious faults taken in the hypervisor that are
due to writable pagetable logic.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 24 Mar 2006 09:59:31 +0000 (10:59 +0100)]
This patch is mainly a cleanup of vmx related xentrace code. One minor
xentrace bug is fixed.
Signed-off-by: Xiaowei Yang <xiaowei.yang@intel.com>
Signed-off-by: Yunfeng Zhao <yunfeng.zhao@intel.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 24 Mar 2006 09:57:52 +0000 (10:57 +0100)]
Update gdb documentation
* Kernel isn't 2.6.12 anymore
* ARCH=xen isn't needed for make anymore
Signed-Off-By: Horms <horms@verge.net.au>
kaf24@firebug.cl.cam.ac.uk [Fri, 24 Mar 2006 09:52:10 +0000 (10:52 +0100)]
The patch removes old Xenbus files, fixes 0 length mmu_updates table bug
(bugfix by Melvin Anderson) and adds missing console.h header file
(again spotted by Melvin).
Signed-off-by: Grzegorz Milos <gm281@cam.ac.uk>
kaf24@firebug.cl.cam.ac.uk [Fri, 24 Mar 2006 09:47:48 +0000 (10:47 +0100)]
Fix mini-os xenbus.
From: Grzegorz Milos <gm281@cam.ac.uk>
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 23 Mar 2006 17:57:21 +0000 (18:57 +0100)]
Remove unnecessary zlib includes.
Signed-off-by: Hollis Blanchard <hollisb@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 23 Mar 2006 17:47:43 +0000 (18:47 +0100)]
on_selected_cpus() must not send IPIs with empty target masks.
This causes send accept errors on Pentium/P6 .
Signed-off-by: Keir Fraser <keir@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 16:37:37 +0000 (17:37 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 16:27:48 +0000 (17:27 +0100)]
Remove undefined XFAIL_TESTS variable.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 16:25:55 +0000 (17:25 +0100)]
Remove undefined XFAIL_TESTS variable.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
cl349@firebug.cl.cam.ac.uk [Thu, 23 Mar 2006 16:21:05 +0000 (16:21 +0000)]
Remove unused variable.
Signed-off-by: Christian Limpach <Christian.Limpach@cl.cam.ac.uk>
kaf24@firebug.cl.cam.ac.uk [Thu, 23 Mar 2006 16:17:08 +0000 (17:17 +0100)]
Reduce spin-waiting in Xen serial driver:
1. Split the serial port lock into receiver and transmitter locks.
2. In the ns16550 interrupt, only call the generic serial service
routines if there is receive (or transmit) work to do.
3. In the generic transmit ISR, avoid long spin-waits by *trying*
to take the transmitter lock and, if that fails, check again
whether the transmitter is empty. This will allow us to bail
bail quickly if there is a long-term lock holder stuffing lots
of bytes.
Also, gdbstub should be setting its serial handle in synchronous mode,
just for sanity.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 23 Mar 2006 14:53:52 +0000 (15:53 +0100)]
Merge hvm_store_cpu_guest_regs() and hvm_store_cpu_guest_ctrl_regs()
into a single function.
On VMX, make the function work even when the passed VCPU is not the
currently-executing VCPU.
This allows gdbserver-xen to get correct cpu register context for VMX guests.
Signed-off-by: Keir Fraser <keir@xensource.com>
Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 23 Mar 2006 14:30:00 +0000 (15:30 +0100)]
Rename generic_page_range as apply_to_page_range, export the symbol
to GPL modules. Rename pte_page parameter to pmd_page (following the
Linux idiom). apply_to_pte_range() should take the pte spinlock.
Signed-off-by: Keir Fraser <keir@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 13:45:21 +0000 (14:45 +0100)]
Merged.
kaf24@firebug.cl.cam.ac.uk [Thu, 23 Mar 2006 13:44:51 +0000 (14:44 +0100)]
Fix 2 VMX time-related bugs:
1) bogomips=0 or a very huge number after booting ia32/ia32e vmx
2) Repeated keys show in Xwindow terminal after one key is pressed
Signed-off-by: Xiaowei Yang <xiaowei.yang@intel.com>
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 13:44:11 +0000 (14:44 +0100)]
Added copyright declaration for Christopher Clark, Andrew Tridgell.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 13:34:35 +0000 (14:34 +0100)]
Use *args inside FAIL and SKIP so that it is possible to call these functions
with multiple arguments, printf style. This is already being done in some
cases (causing an exception of course).
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 13:32:15 +0000 (14:32 +0100)]
Import the current version of talloc from the Samba 3 source base. This gives
us greater confidence that our talloc implementation is "known good". Remove
the OOM handling from consider_message: talloc_set_fail_handler is no longer
supported.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 13:26:38 +0000 (14:26 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 11:59:50 +0000 (12:59 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Thu, 23 Mar 2006 11:59:43 +0000 (12:59 +0100)]
Improve error handling, in particular fixing the ProtocolError that is thrown
when a domain is specified by the user that does not exist. Added a few
error codes -- many more to come, I expect.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Thu, 23 Mar 2006 11:51:10 +0000 (12:51 +0100)]
Enable Windows PAE guest on x86-64. Deals with two issues:
1) To handle the situation when PE=1 and PG=0.
2) Windows PAE guest seems to enable PAE first and then enable PG.
Signed-off-by: Xiaohui Xin <xiaohui.xin@intel.com>