similar to: [PATCH][discuss] evtchn race condition

Index: 2006-10-04/xen/arch/x86/domain.c =================================================================== --- 2006-10-04.orig/xen/arch/x86/domain.c 2006-10-04 15:03:07.000000000 +0200 +++ 2006-10-04/xen/arch/x86/domain.c 2006-10-04 15:11:03.000000000 +0200 @@ -550,10 +550,10 @@ static void load_segments(struct vcpu *n /* CS longword also contains full evtchn_upcall_mask. */

This patch cleans up some 80-odd warnings I was seeing when compiling the xenpci driver, all completely trivial. It also replaces the use of asm for calling cpuid with the __cpuid compiler intrinsic, and modifies calling code accordingly. I''m pretty sure my mailer going to mangle the inline patch (below) so I''m attaching the pristine version. (will try to fix for next time) --

# This is a BitKeeper generated diff -Nru style patch. # # ChangeSet # 2005/03/23 18:27:49-08:00 kmacy@shemp.lab.netapp.com # only bump tmpindex by NKPT-1 once # release sched lock on upcall # Signed-off-by: Kip Macy <kmacy@fsmware.com> # # freebsd-5.3-xen-sparse/i386-xen/i386-xen/machdep.c # 2005/03/23 18:24:47-08:00 kmacy@shemp.lab.netapp.com +0 -4 # remove extra bump of

Remove x86 dependency in drivers/xen/events.c for ia64/xen support introducing include/asm/xen/events.h. Introduce xen_irqs_disabled() to hide regs->flags Introduce xen_do_IRQ() to hide regs->orig_ax. make enum ipi_vector definition arch specific. ia64/xen needs four vectors. Add one rmb() because on ia64 xchg() isn't barrier. Signed-off-by: Isaku Yamahata <yamahata at

Remove x86 dependency in drivers/xen/events.c for ia64/xen support introducing include/asm/xen/events.h. Introduce xen_irqs_disabled() to hide regs->flags Introduce xen_do_IRQ() to hide regs->orig_ax. make enum ipi_vector definition arch specific. ia64/xen needs four vectors. Add one rmb() because on ia64 xchg() isn't barrier. Signed-off-by: Isaku Yamahata <yamahata at

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 <samuel.thibault at eu.citrix.com> Signed-off-by: Isaku Yamahata <yamahata at valinux.co.jp> --- drivers/xen/events.c | 2 +- 1 files changed, 1 insertions(+), 1

This patch adds a facility to block on a particular event channel, for when a domain needs to block, but cannot enable interrupts. A domain uses the call something like this: while (1) { clear_evtchn_pending(evtchn); if (check_for_data(evtchn)) break; HYPERVISOR_block_on(evtchn); } The clear of the pending is needed to ensure that any subsequent calls to block_on() don''t

Remove visibility of the FFXSR CPUID bit to an HVM guest. This patch allows HVM Windows x64 to install/boot on AMD-V platforms. This patches applies cleanly to xen-unstable 13743. Please apply to xen-unstable/3.0.5. If possible, pls apply to xen-3.0.4-testing. --Tom thomas.woller@amd.com AMD Corporation 5204 E. Ben White Blvd. UBC1 Austin, Texas 78741 +1-512-602-0059

Keir, The following 3 patches allow HVM (SVM and VT) guests to be passively profiled using the very latest patches from Renato. These patches apply to 11470. Renato''s patches: http://xenoprof/sourceforge.net oprofile-0.9.1-xen-r2.patch And also a patch not posted FAIK (oprofile-0.9.1-fix-hvm-addr-overlap.patch) which is in the attached tar.bz2 file. hvm_xenoprofile_1.patch - The first

Attached patch for svm will enable a tlb flush for each vmrun. Applies cleanly to 8781:dcc2beb8a1d2 Signed-off-by: Tom Woller <thomas.woller@amd.com> _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

Attached simple patch modifies the rombios and hvmloader Makefiles to default to an 8-way bochs bios for the hvmloader. Is there a reason to not define the mp config table by default for >1 processors (8), in the repository? Seems like using the hvm guest config file can the methodolody to describe a UP guest regardless of the mptables. Defaulting the rombios to MP would allow users to

Patch 2/2 - Add flag to indicate that an exception event needs injecting, and to delay the ext interrupt injection. Remove unnecessary check of RFLAGS.IF for ExtInt injection. Applies cleanly to xen-unstable c/s 11831. Please apply to xen-unstable.hg. We would also want this patch to be in a 3.0.3-1 base whenever that is branched. Signed-off-by: Travis Betak <travis.betak@amd.com>

Compilation failure on c/s 17194 is as follows: /root/randy/vtd-stage/xen/common/built_in.o: In function `guest_remove_page'': /root/randy/vtd-stage/xen/common/memory.c:172: undefined reference to `__bitop_bad_size'' /root/randy/vtd-stage/xen/common/memory.c:172: relocation truncated to fit: R_X86_64_PC32 against undefined symbol `__bitop_bad_size''

Those two should always be delivered first imo. Signed-off-by: Jan Beulich <jbeulich@suse.com> --- a/xen/arch/x86/hvm/irq.c +++ b/xen/arch/x86/hvm/irq.c @@ -395,16 +395,16 @@ struct hvm_intack hvm_vcpu_has_pending_i struct hvm_domain *plat = &v->domain->arch.hvm_domain; int vector; - if ( (plat->irq.callback_via_type == HVMIRQ_callback_vector) -

This set of patches adds the support for acceleration plugins to the netfront/netback drivers. These plugins are intended to support virtualisable network hardware that can be directly accessed from the guest, bypassing dom0. This is in response to the RFC we posted to xen-devel with an outline of our approach at the end of September. To follow will be another set of patches to provide our

Doing some janitorial (you cleaning the flooded toilets and such) work today. I have come across this line of code that really I''m not sure what the intent was..in xen/include/sched.h #define hypercall_preempt_check() (unlikely( \ softirq_pending(smp_processor_id()) | \ (!!current->vcpu_info->evtchn_upcall_pending & \

Hi, I''m trying to boot up with lot more than 32 vcpus on this very large box. I overcame vcpu_info[MAX_VIRT_CPUS] by doing vcpu placement hypercall in guest, but now running into evt channel limit (lots of devices): unsigned long evtchn_pending[sizeof(unsigned long) * 8]; which limits to 512 max for my 64bit dom0. The only recourse seems to create a new struct shared_info_v2{},

Is it intentional that - under SVM, 32-bit guests can freely set EFER.LME - under VMX, 32-bit guests can''t access EFER at all? Thanks, Jan _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel

Renato, If one creates an HVM domain (say domain 1) and then does echo 1 > /dev/oprofile/passive_domains one then gets the following on the Xen serial console: (XEN) xenoprof.c:143:d0 xenoprof/x86 with autotranslated mode enabledisn''t supported yet (XEN) xenoprof.c:143:d0 xenoprof/x86 with autotranslated mode enabledisn''t supported yet (XEN) xenoprof.c:143:d0

I am currently looking at a bug report[1] which is happening when a Xen PVM guest with multiple VCPUs is running a high IO database load (a test script is available in the bug report). In experimenting it seems that this happens (or is getting more likely) when the number of VCPUs is 8 or higher (though I have not tried 6, only 2 and 4), having autogroup enabled seems to make it more likely, too