emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 14:15:02 +0000 (15:15 +0100)]
Increase the static constant default number of loopback interfaces, nloopbacks,
to 8. These interfaces are very cheap, so there's no problem having a few
spare. Anyone with more than 8 NICs can still use the kernel command line or
the module parameter line to pass a larger limit to this module.
Detail the kernel command line parameter for nloopbacks in the diagnostic in
network-bridge.
Closes bug #381.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 13:56:14 +0000 (14:56 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 13:53:27 +0000 (14:53 +0100)]
Fix two leaks in the pyxc interface. The dictionaries are being added
to lists, and an extra ref was being left behind.
Signed-off-by: Sean Perry <sean@xensource.com>
Signed-off-by: Robert Read <robert@xensource.com>
Michael.Fetterman@cl.cam.ac.uk [Tue, 8 Nov 2005 13:41:25 +0000 (14:41 +0100)]
merge
Michael.Fetterman@cl.cam.ac.uk [Tue, 8 Nov 2005 12:30:05 +0000 (13:30 +0100)]
merge
Michael.Fetterman@cl.cam.ac.uk [Tue, 8 Nov 2005 12:26:50 +0000 (13:26 +0100)]
Cleanup various shadow mode asserts.
Separate out the ability for domains to be able to write to their
pagetables (ala "writable page tables", which uses write-protected PTEs
to address the page tables: this is shadow_mode_write_all()) from the
right of a domain to create a PTE with write permissions that points
at a page table (this is shadow_mode_wr_pt_pte())...
Minor cleanup of SHADOW_DEBUG (at least make it compilable) in shadow.c.
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 11:31:43 +0000 (12:31 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 11:30:38 +0000 (12:30 +0100)]
Added file missed on commit of earlier patch by Dan Smith.
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 11:28:19 +0000 (12:28 +0100)]
This patch removes all of the auto-generated files that were checked
in with xm-test in changeset 7634. It also adds an "autogen" script
that re-creates these files.
The auto-generated files really don't need to be in the tree, as they
should be generated by the user on the machine where they plan to run
xm-test.
Signed-off-by: Dan Smith <danms@us.ibm.com>
mafetter@aosda.research.intel-research.net [Tue, 8 Nov 2005 11:26:48 +0000 (12:26 +0100)]
Bug fix in error handling in get_page_from_l4e()
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Nov 2005 11:25:09 +0000 (12:25 +0100)]
Remove unused and unnecessary gnttab ioctl.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Nov 2005 11:17:53 +0000 (12:17 +0100)]
Update tools/examples docs.
Signed-off-by: Nivedita Singhvi <niv@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Nov 2005 11:16:37 +0000 (12:16 +0100)]
After move randomMac to netif in 7637, we need to this for vmx guest.
otherwise, if no mac address configured, vmx guest can not be created.
Signed-off-by: Xiaofeng Ling <xiaofeng.ling@intel.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Nov 2005 11:14:33 +0000 (12:14 +0100)]
Remove the execute bit from shell snipplets (the -common files not
starting with #!) found in the example directory under tools. It
further changes the installation path of the udev rules to comply with
what debian does: It places the rules file in /etc/udev/ und creates a
symlink to it in /etc/udev/rules.d.
Signed-off-by: Tobias Hunger <tobias@aquzul.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Nov 2005 11:06:18 +0000 (12:06 +0100)]
Fix the clean target of the tools/xenstore Makefile to
remove all files created. Currently leaves some of its
files behind.
Signed-off-by: Tobias Hunger <tobias@aquzul.com>
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 11:02:10 +0000 (12:02 +0100)]
This patch increments the version number of xm-test so that we can
distinguish reports from the 0.4.0 tarball and in-tree versions.
Signed-off-by: Dan Smith <danms@us.ibm.com>
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 11:01:09 +0000 (12:01 +0100)]
This patch adds a URL to the reporting site, as well as a note about
running with lots of free memory to the README.
Signed-off-by: Dan Smith <danms@us.ibm.com>
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 11:00:23 +0000 (12:00 +0100)]
1. Set the UID and EUID of the process when trying to switch to a
non-root user.
2. Better handle the debug-printing inside the console
Signed-off-by: Dan Smith <danms@us.ibm.com>
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 10:56:55 +0000 (11:56 +0100)]
1. Unmark the 07 and 08 tests of block_attach as XFAIL
2. Change the MEM_PER_DOM amount in 11_create_concurrent to 24MB, so
that the test will run on x86_64.
3. Increase the number of DomUs created in 12_create_concurrent to 5
4. Fix a logic error in 01_unpause
5. Add a test 06_help that checks all xm commands with 0-10 arguments
to make sure none of them traceback with varying argument counts.
Signed-off-by: Dan Smith <danms@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 8 Nov 2005 10:44:48 +0000 (11:44 +0100)]
merge
mafetter@seana.research.intel-research.net [Tue, 8 Nov 2005 10:38:32 +0000 (11:38 +0100)]
Fixed domain refcounting bug (found by inspection)
mafetter@seana.research.intel-research.net [Tue, 8 Nov 2005 10:36:55 +0000 (11:36 +0100)]
Fixed a bug in the audit code.
We don't hold a type refcount to the user's ptbase frame in
shadow refcounting mode.
unknown@aosda.research.intel-research.net [Tue, 8 Nov 2005 10:36:37 +0000 (11:36 +0100)]
Fixed domain refcounting bug (found by inspection)
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 10:35:25 +0000 (11:35 +0100)]
Remove unused parameter.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
mafetter@seana.research.intel-research.net [Tue, 8 Nov 2005 10:31:16 +0000 (11:31 +0100)]
Fix vncviewer spawning so it actually works.
How did this ever work for other people?
mafetter@seana.research.intel-research.net [Tue, 8 Nov 2005 10:29:07 +0000 (11:29 +0100)]
Added some minor defensive coding/error handling in xc_linux_build.c
djm@kirby.fc.hp.com [Tue, 8 Nov 2005 02:40:31 +0000 (20:40 -0600)]
Merge xen-unstable into xen-ia64-unstable again (to get ioapic changes)
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 02:35:25 +0000 (03:35 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 02:34:23 +0000 (03:34 +0100)]
Push the process of waiting for devices to come up right out of DevController,
and drive this process from xm create. Architecturally, this separates the
process of domain creation from the process of waiting for devices, allowing
tools to choose when to perform that wait (if at all). This places the waiting
in the same category as the unpause after domain creation, architecturally.
The main advantage to this approach is that it takes waiting for devices out
of the scope of XendDomain's domains_lock. When restarting a domain, the
watch would fire for @releaseDomain while we were waiting for new domains to
come up. This would deadlock the watch thread, as no more new watches could
be delivered. Closes bug #387.
In the longer term, we expect to be able to wait for devices to come up
completely, not simply to wait for the hotplug scripts to run, and so it is
necessary then to move the waiting procedure right out of the server, so that
it can be performed after the domain unpause. Without unpausing the domain,
the frontends will not come up, and so we cannot detect successful device
completion.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 02:07:57 +0000 (03:07 +0100)]
Rearrange restart logic so that a failure to restart a domain does not cause
cascading secondary errors. Fixes bug #384.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 01:51:34 +0000 (02:51 +0100)]
Mark plenty of variables as unused, and remove an unused import.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Tue, 8 Nov 2005 00:26:16 +0000 (01:26 +0100)]
Remove unused XENSTORED_XYZ constants.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
djm@kirby.fc.hp.com [Mon, 7 Nov 2005 22:53:25 +0000 (16:53 -0600)]
Enable ar.unat handling for fast paths (by Anthony Xu)
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 20:22:02 +0000 (21:22 +0100)]
Rename parameter to avoid shadowing builtin.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
unknown@aosda.research.intel-research.net [Mon, 7 Nov 2005 17:35:55 +0000 (18:35 +0100)]
Minor Makefile cleanup in tools/libxc/Makefile
djm@kirby.fc.hp.com [Mon, 7 Nov 2005 17:25:59 +0000 (11:25 -0600)]
Missed ia64 changes to one drivers/xen file
unknown@aosda.research.intel-research.net [Mon, 7 Nov 2005 17:25:09 +0000 (18:25 +0100)]
Make xen0's default config match itself after a "make oldconfig"
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Nov 2005 17:14:45 +0000 (18:14 +0100)]
Fix pagetable pinning logic for xen/i386 kernels. The pin
flag is now associated with the pgd rather than the mm -- this
avoids a race where a pgd is allocated from the pgd_cache but,
before it gets associated with an mm, the kernel suspends itself.
At this point the kernel mappings will not get rewritten when the
kernel is resumed, and the system will fail.
A further advantage is that the code is slightly simpler and less
invasive (no changes to mm_context for example).
Signed-off-by: Keir Fraser <keir@xensource.com>
djm@kirby.fc.hp.com [Mon, 7 Nov 2005 17:13:38 +0000 (11:13 -0600)]
Merge latest xen-unstable into xen-ia64-unstable
unknown@aosda.research.intel-research.net [Mon, 7 Nov 2005 16:54:55 +0000 (17:54 +0100)]
Minor updates to .hgignore
Added "tools/.*/TAGS" entry.
Sorted all entries.
djm@kirby.fc.hp.com [Mon, 7 Nov 2005 16:40:42 +0000 (10:40 -0600)]
Fast tlb miss reflection (mostly working, missing robustness, default off)
Signed-off by: Dan Magenheimer <dan.magenheimer@hp.com>
I've been maintaining this out-of-tree partially finished for some time
so am now checking it in but default off. The current version needs
some additional error checking and only works (fast) for domain0, but
it makes a significant performance differenece.
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Nov 2005 15:37:58 +0000 (16:37 +0100)]
This patch remove the obsolete ioapic code on qemu device model side.
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Nov 2005 15:36:27 +0000 (16:36 +0100)]
This patch add virtual IOAPIC support for VMX guest.
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Xin Li <xin.b.li@intel.com>
Signed-off-by: Eddie Dong <eddie.dong@intel.com>
Signed-off-by: Jun Nakajima <jun.nakajima@intel.com>
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Nov 2005 15:35:46 +0000 (16:35 +0100)]
This patch change the vmx mmio handler to static.
Signed-off-by: Yunhong Jiang <yunhong.jiang@intel.com>
Signed-off-by: Xin Li <xin.b.li@intel.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 14:06:27 +0000 (15:06 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 14:06:00 +0000 (15:06 +0100)]
Added description of the behaviour of rename-restart.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
jbulpin@wicket.uk.xensource.com [Mon, 7 Nov 2005 12:04:27 +0000 (13:04 +0100)]
Fixed syntax error in iopif patch.
Signed-off-by: James Bulpin <james@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Nov 2005 11:53:38 +0000 (12:53 +0100)]
merge
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Nov 2005 11:52:49 +0000 (12:52 +0100)]
fast_flush_area() should index the unmap request array using
invcount, not i. Also, fix initialisation of pending_grant_handles:
it was happening too late, and the use of memset() was bogus.
Signed-off-by: Harry Butterworth <butterwo@uk.ibm.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 11:36:53 +0000 (12:36 +0100)]
Move the randomMAC method from xm.create into server.netif. This way, it can
be shared with other command line tools, and the xm network-attach command.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 11:36:44 +0000 (12:36 +0100)]
Fix misnamed call to legacy_mask_to_prefix. Thanks to Tim Moreton for finding
this.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 11:36:35 +0000 (12:36 +0100)]
Diagnostic for hitting the nloopbacks limit. This is presumably the cause of
all the "two bridges don't work" bugs that have been reported (bug #381).
Consideration will be given to raising the default limit -- this is just a
diagnostic of the problem.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 11:36:16 +0000 (12:36 +0100)]
Wholesale import of xm-test 0.4.0.
This is not yet integrated with the rest of the Xen build system.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 11:35:06 +0000 (12:35 +0100)]
Hack around problem whereby VMX domains (using the ioemu device) now fail
because the new hotplug script error checking kicks in.
ioemu really ought to move out of blkif.py.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 11:30:35 +0000 (12:30 +0100)]
Added assertion, to stop it blowing up more obscurely a little later.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 11:30:16 +0000 (12:30 +0100)]
Tidy exception handling on IntroduceDomain, quieten down one debug message, add
another.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
emellor@leeni.uk.xensource.com [Mon, 7 Nov 2005 11:28:53 +0000 (12:28 +0100)]
Make xenstored reopen its trace file on SIGHUP. This allows one to rotate the
logs beneath the daemon.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Mon, 7 Nov 2005 11:05:15 +0000 (12:05 +0100)]
Add support for the ioport_permission dom0 op to xend and xm
xm now accepts a parameter 'ioports' that accepts a hex ioport
or ioport range, in the form 02f8[-02ff]
Signed-off-by: Jody Belka <knew (at) pimb (dot) org>
kaf24@firebug.cl.cam.ac.uk [Sun, 6 Nov 2005 18:50:33 +0000 (19:50 +0100)]
Revert c/s 7626 -- it's missing a new file.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Sun, 6 Nov 2005 16:50:24 +0000 (17:50 +0100)]
Fix error path in blkback driver where a grant mapping in the
middle of a batch fails. Spotted by Harry Butterworth (IBM).
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Sun, 6 Nov 2005 15:40:33 +0000 (16:40 +0100)]
Add support for the ioport_permission dom0 op to xend and xm
xm now accepts a parameter 'ioports' that accepts a hex ioport
or ioport range, in the form 02f8[-02ff]
Signed-off-by: Jody Belka <knew (at) pimb (dot) org>
kaf24@firebug.cl.cam.ac.uk [Sun, 6 Nov 2005 15:40:15 +0000 (16:40 +0100)]
Added domain_ioport_permission to the python module xen.lowlevel.xc,
wrapping the equivalent libxc call
Signed-off-by: Jody Belka <knew (at) pimb (dot) org>
kaf24@firebug.cl.cam.ac.uk [Sun, 6 Nov 2005 15:39:55 +0000 (16:39 +0100)]
Added xc_domain_ioport_permission to libxc, which wraps the dom0 op
Signed-off-by: Jody Belka <knew (at) pimb (dot) org>
kaf24@firebug.cl.cam.ac.uk [Sun, 6 Nov 2005 15:39:02 +0000 (16:39 +0100)]
This patch adds a new boot parameter, dom0_ioports_disable, which
accepts a comma seperated list of hex ioports and/or ioport ranges
(eg. dom0_ioports_disable=02f8-02ff,03f8-03ff), and applies them to dom0.
Signed-off-by: Jody Belka <knew (at) pimb (dot) org>
kaf24@firebug.cl.cam.ac.uk [Sat, 5 Nov 2005 10:30:01 +0000 (11:30 +0100)]
This patch is intended to make qemu support ia64/vti. We have validated
it against latest xen-unstable.hg, both ia32 and x86-64 are not
affected.
Signed-off-by Ke Yu <ke.yu@intel.com>
Signed-off-by Kevin Tian <kevin.tian@intel.com>
Signed-off-by Nakajima Jun <nakajima.jun@intel.com>
Signed-off-by Anthony Xu <anthony.xu@intel.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 5 Nov 2005 10:26:29 +0000 (11:26 +0100)]
This patch is to remove the pit_timer when the vmx domain is
inactive to save HV external IRQ caused by ac_time and some cleanup for
ioapic in HV.
Thx,eddie
Signed-off-by: Eddie Dong <eddie.dong@intel.com>
kaf24@firebug.cl.cam.ac.uk [Sat, 5 Nov 2005 10:23:02 +0000 (11:23 +0100)]
vlapic_init alloc memory for struct vlapic, so it should do a memset on
it, not vlapic_reset.
Signed-off-by: Xin Li <xin.b.li@intel.com>
smh22@firebug.cl.cam.ac.uk [Sat, 5 Nov 2005 09:55:38 +0000 (10:55 +0100)]
Quiet down balloon driver.
Signed-off-by: Steven Hand <steven@xensource.com>
smh22@firebug.cl.cam.ac.uk [Sat, 5 Nov 2005 09:54:12 +0000 (10:54 +0100)]
Check correct flag in grant transfer - this eliminates the most common
source of 'zombie-' domains.
Signed-off-by: Steven Hand <steven@xensource.com>
djm@kirby.fc.hp.com [Fri, 4 Nov 2005 16:40:29 +0000 (10:40 -0600)]
Minor driver code changes/rearrangement to reduce ia64-specific patches
Signed-off by: Dan Magenheimer <dan.magenheimer@hp.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 4 Nov 2005 15:01:49 +0000 (16:01 +0100)]
Remove shadow audit debugging. Too noisy for
log-dirty mode.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 4 Nov 2005 10:46:10 +0000 (11:46 +0100)]
Specifying 'mem=' as Linux boot parameter disables the default
8MB slack in the mem_map and the p2m table.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 4 Nov 2005 10:08:45 +0000 (11:08 +0100)]
This is a fix for some device model corner cases including:
1: mistake sequence of set/clr to virtual IRQ line.
2: wrong sequence of IRQ request clear
3: Add one more place to enable_irq_window.
Signed-off-by: Eddie Dong <eddie.dong@intel.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 4 Nov 2005 10:06:41 +0000 (11:06 +0100)]
This patch adds a check to make sure the config file passed to
vnet-create is correct. This is better than throwing IOError, because
IOError's are currently reported as xend communication failures.
Also, fix the IOError handling at the end of the file so we don't get
a stack dump if that path is taken.
Signed-off-by: Dan Smith <danms@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 4 Nov 2005 10:05:10 +0000 (11:05 +0100)]
This is another patch that adds an arg_check() to domid and domname so
that we don't dump stack if a user runs either without an argument.
Signed-off-by: Dan Smith <danms@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Fri, 4 Nov 2005 10:03:20 +0000 (11:03 +0100)]
This patch adds an arg_check into the set_vcpus code in xm, so that we
don't output a stack trace if <2 arguments are passed.
Signed-off-by: Dan Smith <danms@us.ibm.com>
djm@kirby.fc.hp.com [Fri, 4 Nov 2005 03:37:29 +0000 (21:37 -0600)]
Enable test for running_on_xen from modules
Signed-off by: Dan Magenheimer <dan.magenheimer@hp.com>
djm@kirby.fc.hp.com [Fri, 4 Nov 2005 00:06:15 +0000 (18:06 -0600)]
Fix gcc-3.4 compiler errors (submitted by Takebe Akio)
djm@kirby.fc.hp.com [Thu, 3 Nov 2005 22:50:44 +0000 (16:50 -0600)]
Oops, backout accidental test config, committed in wrong directory
djm@kirby.fc.hp.com [Thu, 3 Nov 2005 22:01:38 +0000 (16:01 -0600)]
Test new config
djm@kirby.fc.hp.com [Thu, 3 Nov 2005 13:21:26 +0000 (07:21 -0600)]
Fix console output and bug in evtchn_ia64 (by Tristan Gingold)
djm@kirby.fc.hp.com [Thu, 3 Nov 2005 13:14:09 +0000 (07:14 -0600)]
Fix ar.unat save/restore for fast rfi (by Anthony Xu)
kaf24@firebug.cl.cam.ac.uk [Thu, 3 Nov 2005 08:58:45 +0000 (09:58 +0100)]
Simplified vmx_platform.c by removing obsolete code and redundant vmread's.
Signed-Off-By: Leendert van Doorn <leendert@watson.ibm.com>
emellor@leeni.uk.xensource.com [Thu, 3 Nov 2005 01:45:07 +0000 (02:45 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Thu, 3 Nov 2005 01:44:41 +0000 (02:44 +0100)]
Fix handling of child in forkHelper.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 2 Nov 2005 18:12:30 +0000 (19:12 +0100)]
Fix map_domain_page() by reverting a bogus simplification
that I committed in changeset 5303:
d68bc64e2cfe38a85483a8c5d349b4ee45e91fb9
This fixes some very weird and nasty Xen crashes.
Signed-off-by: Keir Fraser <keir@xensource.com>
djm@kirby.fc.hp.com [Wed, 2 Nov 2005 17:45:49 +0000 (11:45 -0600)]
Adapt for recent change in xen-unstable common/trace.c
sos22@douglas.cl.cam.ac.uk [Wed, 2 Nov 2005 16:29:32 +0000 (17:29 +0100)]
Merge.
Signed-off-by: Steven Smith, sos22@cam.ac.uk
emellor@leeni.uk.xensource.com [Wed, 2 Nov 2005 15:43:32 +0000 (16:43 +0100)]
Merged.
emellor@leeni.uk.xensource.com [Wed, 2 Nov 2005 15:42:29 +0000 (16:42 +0100)]
Rewritten handling of child stderr to avoid deadlock.
Signed-off-by: Ewan Mellor <ewan@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 2 Nov 2005 10:25:21 +0000 (11:25 +0100)]
Make the -V option of xenconsoled actually do something.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 2 Nov 2005 10:25:08 +0000 (11:25 +0100)]
Make it clear that the result of xs_get_domain_path() must be free()'d.
kaf24@firebug.cl.cam.ac.uk [Wed, 2 Nov 2005 10:24:49 +0000 (11:24 +0100)]
Fix memory leak in domain cleanup spotted by doremifa in #xen.
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 2 Nov 2005 10:21:46 +0000 (11:21 +0100)]
The attached patch to 1) fix a 64-bit bug with some of the data packing,
and 2) adds a workaround for an Infineon TPM hardware bug.
Signed-off-by: Vinnie Scarlata, Joseph Cihula (@intel.com)
kaf24@firebug.cl.cam.ac.uk [Wed, 2 Nov 2005 10:19:48 +0000 (11:19 +0100)]
fix VMX decoder for MOVZ instruction.
In VMX decoder, we always assume the 2 operands of x86 instructions are
in the same length, however, MOVZ is an exception.
This patch fixes MOVZ instruction handling.
Signed-off-by: Xin Li <xin.b.li@intel.com>
Signed-off-by: Dan Xu <dan.d.xu@intel.com>
kaf24@firebug.cl.cam.ac.uk [Wed, 2 Nov 2005 10:18:51 +0000 (11:18 +0100)]
Just some small code cleanup to shadow.c, no logic changed.
Signed-off-by: Xin Li <xin.b.li@intel.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 1 Nov 2005 23:08:01 +0000 (00:08 +0100)]
Changeset 7419 introduced a check for DOM0 before setting a bit in the
cpu_present map. This prevents domU SMP kernels without HOTPLUG_CPU
support from booting. Secondary cpus need to be present and online
before init/main.c:do_basic_setup() calls init_workqueues() (which
initializes per-cpu workqueues).
Without this patch, non HOTPLUG_CPU enabled kernels hang when flushing
cpu workqueues as the spinlock in the structure is never initialized (it
has a default value of zero which means the lock has been acquired on
x86).
Signed-off-by: Ryan Harper <ryanh@us.ibm.com>
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 1 Nov 2005 18:13:06 +0000 (19:13 +0100)]
>I'm hitting an ASSERT in alloc_monitor_pagetable() with debug=y and I
>can reproduce the condition when debug isn't on. The monitor_table is
>not zeroed when entering alloc_monitor_pagetable() - this happens when I
>create a vmx guest. I'm not sure why this happening at first glance.
>File: xen/arch/x86/shadow32.c
>Line: 689
>System: VT/VMX FC3 x86_32 (I can reproduce on FC4 x86_64 too)
>ASSERT(pagetable_get_paddr(v->arch.monitor_table) == 0);
Solution:
What happens is that the monitor page table is not zeroed out
during the initial setup for VMX guests. As a result, at a later
time when we enable shadow mode for these VMX guests, we trip on
the ASSERT problem above. To fix this, we just need to initialize
the monitor page table to 0 before calling vmx_final_setup_guest()
and enabling shadow mode.
kaf24@firebug.cl.cam.ac.uk [Tue, 1 Nov 2005 16:49:19 +0000 (17:49 +0100)]
Fix 'nosmp' and ensure that CPU#0 is reported properly
by /proc/cpuinfo. Also fix the sibling and core maps.
Signed-off-by: Keir Fraser <keir@xensource.com>
kaf24@firebug.cl.cam.ac.uk [Tue, 1 Nov 2005 15:44:25 +0000 (16:44 +0100)]
Temporary fix for blkfront failure on save/restore.
Signed-off-by: Keir Fraser <keir@xensource.com>