- 06 5月, 2012 13 次提交
-
-
由 Benjamin Herrenschmidt 提交于
There is nothing in the code for emulating TCE tables in the kernel that prevents it from working on "PR" KVM... other than ifdef's and location of the code. This and moves the bulk of the code there to a new file called book3s_64_vio.c. This speeds things up a bit on my G5. Signed-off-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org> [agraf: fix for hv kvm, 32bit, whitespace] Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Mihai Caraman 提交于
Guest r8 register is held in the scratch register and stored correctly, so remove the instruction that clobbers it. Guest r13 was missing from vcpu, store it there. Signed-off-by: NMihai Caraman <mihai.caraman@freescale.com> Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Alexander Graf 提交于
While handling an exit, we should listen for interrupts and make sure to receive them when they arrive, to keep our latencies low. Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Alexander Graf 提交于
When running on a system that is HV capable, some interrupts use HSRR SPRs instead of the normal SRR SPRs. These are also used in the Linux handlers to jump back to code after an interrupt got processed. Unfortunately, in our "jump back to the real host handler after we've done the context switch" code, we were only setting the SRR SPRs, rendering Linux to jump back to some invalid IP after it's processed the interrupt. This fixes random crashes on p7 opal mode with PR KVM for me. Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Alexander Graf 提交于
Stbux writes the address it's operating on to the register specified in ra, not into the data source register. Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Mihai Caraman 提交于
Interrupt code used PPC_LL/PPC_STL macros to load/store some of u32 fields which led to memory overflow on 64-bit. Use lwz/stw instead. Signed-off-by: NMihai Caraman <mihai.caraman@freescale.com> Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Alexander Graf 提交于
While messing around with the SLBs we're running in real mode. The entry to guest space goes through rfid, which is context synchronizing, so there's no need to manually synchronize anything through isync. With this patch and a simple priviledged SPR access loop guest, I get a speed bump from 2035607 to 2181301 exits per second. Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Alexander Graf 提交于
By shuffling a few instructions around we can execute more memory loads in parallel, giving us a small performance boost. With this patch and a simple priviledged SPR access loop guest, I get a speed bump from 2013052 to 2035607 exits per second. Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Varun Sethi 提交于
For Guest accessible SPRGs 4-7, save/restore must be handled differently for 64bit and non-64 bit case. Use the PPC_STD/PPC_LD macros for saving/restoring to/from these registers. Signed-off-by: NVarun Sethi <Varun.Sethi@freescale.com> Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Alexander Graf 提交于
We only want asm code macros to be accessible from asm code, so #ifdef it depending on it. Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Varun Sethi 提交于
Introduced PPC_STD/PPC_LD macros for saving/restoring guest registers to/from their 64 bit copies. Signed-off-by: NVarun Sethi <Varun.Sethi@freescale.com> Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Bharat Bhushan 提交于
Time for which the hrtimer is started for decrementer emulation is calculated using tb_ticks_per_usec. While hrtimer uses the clockevent for DEC reprogramming (if needed) and which calculate timebase ticks using the multiplier and shifter mechanism implemented within clockevent layer. It was observed that this conversion (timebase->time->timebase) are not correct because the mechanism are not consistent. In our setup it adds 2% jitter. With this patch clockevent multiplier and shifter mechanism are used when starting hrtimer for decrementer emulation. Now the jitter is < 0.5%. Signed-off-by: NBharat Bhushan <bharat.bhushan@freescale.com> Signed-off-by: NAlexander Graf <agraf@suse.de>
-
由 Bharat Bhushan 提交于
Keep track of minimum and maximum address mapped by tlb1. This helps in TLBMISS handling in KVM to quick check whether the address lies in mapped range. If address does not lies in this range then no need to look in each tlb1 entry of tlb1 array. Signed-off-by: NBharat Bhushan <bharat.bhushan@freescale.com> Signed-off-by: NAlexander Graf <agraf@suse.de>
-
- 03 5月, 2012 1 次提交
-
-
由 Christian Borntraeger 提交于
Let userspace know the number of max and supported cpus for kvm on s390. Return KVM_MAX_VCPUS (currently 64) for both values. Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
- 01 5月, 2012 3 次提交
-
-
由 Cornelia Huck 提交于
Handle the mandatory intercept SET CLOCK PROGRAMMABLE FIELD instruction. Signed-off-by: NCornelia Huck <cornelia.huck@de.ibm.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
由 Christian Borntraeger 提交于
Lets replace the old open coded version of diag 0x44 (which relied on compat_sched_yield) with kvm_vcpu_on_spin. Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
由 Konstantin Weitz 提交于
This patch implements the directed yield hypercall found on other System z hypervisors. It delegates execution time to the virtual cpu specified in the instruction's parameter. Useful to avoid long spinlock waits in the guest. Christian Borntraeger: moved common code in virt/kvm/ Signed-off-by: NKonstantin Weitz <WEITZKON@de.ibm.com> Signed-off-by: NChristian Borntraeger <borntraeger@de.ibm.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
- 28 4月, 2012 3 次提交
-
-
由 Jan Kiszka 提交于
We can't run PIT IRQ injection work in the interrupt context of the host timer. This would allow the user to influence the handler complexity by asking for a broadcast to a large number of VCPUs. Therefore, this work was pushed into workqueue context in 9d244caf2e. However, this prevents prioritizing the PIT injection over other task as workqueues share kernel threads. This replaces the workqueue with a kthread worker and gives that thread a name in the format "kvm-pit/<owner-process-pid>". That allows to identify and adjust the kthread priority according to the VM process parameters. Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
由 Jan Kiszka 提交于
Add descriptions for KVM_CREATE_PIT2 and KVM_GET/SET_PIT2. Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
由 Jan Kiszka 提交于
This helps to identify sections and it also fixes the numbering from 4.54 to 4.61. Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
- 24 4月, 2012 3 次提交
-
-
由 Avi Kivity 提交于
'bool' wants 8-bit registers. Reported-by: NTakuya Yoshikawa <takuya.yoshikawa@gmail.com> Signed-off-by: NAvi Kivity <avi@redhat.com>
-
由 Gleb Natapov 提交于
The patch introduces a bitmap that will hold reasons apic should be checked during vmexit. This is in a preparation for vp eoi patch that will add one more check on vmexit. With the bitmap we can do if(apic_attention) to check everything simultaneously which will add zero overhead on the fast path. Signed-off-by: NGleb Natapov <gleb@redhat.com> Signed-off-by: NAvi Kivity <avi@redhat.com>
-
由 Jan Kiszka 提交于
Currently, MSI messages can only be injected to in-kernel irqchips by defining a corresponding IRQ route for each message. This is not only unhandy if the MSI messages are generated "on the fly" by user space, IRQ routes are a limited resource that user space has to manage carefully. By providing a direct injection path, we can both avoid using up limited resources and simplify the necessary steps for user land. Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com> Signed-off-by: NAvi Kivity <avi@redhat.com>
-
- 21 4月, 2012 1 次提交
-
-
由 Marcelo Tosatti 提交于
Needed by kvm_para_has_feature(). Reported-by: NStephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
- 20 4月, 2012 5 次提交
-
-
由 Avi Kivity 提交于
s/kcm/kvm/. Signed-off-by: NAvi Kivity <avi@redhat.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
由 Avi Kivity 提交于
MMIO that are split across a page boundary are currently broken - the code does not expect to be aborted by the exit to userspace for the first MMIO fragment. This patch fixes the problem by generalizing the current code for handling 16-byte MMIOs to handle a number of "fragments", and changes the MMIO code to create those fragments. Signed-off-by: NAvi Kivity <avi@redhat.com> Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
由 Marcelo Tosatti 提交于
Merge reason: development work has dependency on kvm patches merged upstream. Conflicts: Documentation/feature-removal-schedule.txt Signed-off-by: NMarcelo Tosatti <mtosatti@redhat.com>
-
git://people.freedesktop.org/~airlied/linux由 Linus Torvalds 提交于
Pull drm fixes from Dave Airlie: "It's like a grab bag of one liners: - core: fix page flip error path, reorder object teardown. - usb: fix the drm_usb module license. - i915: VT switch on SNB with non-native modes fix, and a regression fix from 3.3. - radeon: missing unreserve on SI, AGP/VRAM setup fix (fixes radeon on IA64, but its a generic bug), an rn50 regression from 3.3, turn off MSIs on rv515 (it loses rearms every so often)." * 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: nouveau: Set special lane map for the right chipset drm/radeon: fix load detect on rn50 with hardcoded EDIDs. drm: Releasing FBs before releasing GEM objects during drm_release drm/nouveau/pm: don't read/write beyond end of stack buffer drivers: gpu: drm: gma500: mdfld_dsi_output.h: Remove not unneeded include of version.h radeon: fix r600/agp when vram is after AGP (v3) drm: fix page_flip error handling drm/radeon/kms: fix the regression of DVI connector check drm/usb: fix module license on drm/usb layer. drm/i915: Do not set "Enable Panel Fitter" on SNB pageflips drm/i915: Hold mode_config lock whilst changing mode for lastclose() drm/radeon/si: add missing radeon_bo_unreserve in si_rlc_init() v2 drm/radeon: disable MSI on RV515 drm/i915: don't clobber the special upscaling lvds timings
-
git://git.kernel.org/pub/scm/virt/kvm/kvm由 Linus Torvalds 提交于
Pull KVM updates from Marcelo Tosatti. * git://git.kernel.org/pub/scm/virt/kvm/kvm: KVM: lock slots_lock around device assignment KVM: VMX: Fix kvm_set_shared_msr() called in preemptible context KVM: unmap pages from the iommu when slots are removed KVM: PMU emulation: GLOBAL_CTRL MSR should be enabled on reset
-
- 19 4月, 2012 11 次提交
-
-
由 Henrik Rydberg 提交于
The refactoring of the nv50 logic, introduced in 8663bc7c, modified the test for the special lane map used on some Apple computers with Nvidia chipsets. The tested MBA3,1 would still boot, but resume from suspend stopped working. This patch restores the old test, which fixes the problem. Signed-off-by: NHenrik Rydberg <rydberg@euromail.se> Acked-by: NBen Skeggs <bskeggs@redhat.com> Signed-off-by: NDave Airlie <airlied@redhat.com>
-
由 Dave Airlie 提交于
When the force changes went in back in 3.3.0, we ended up returning disconnected in the !force case, and the connected in when forced, as it hit the hardcoded check. Fix it so all exits go via the hardcoded check and stop spurious modesets on platforms with hardcoded EDIDs. Reported-by: Evan McNabb (Red Hat) Reviewed-by: NAlex Deucher <alexander.deucher@amd.com> Cc: stable@vger.kernel.org Signed-off-by: NDave Airlie <airlied@redhat.com>
-
由 Prathyush 提交于
During DRM release, all the FBs and gem objects are released. If a gem object is being used as a FB and set to a crtc, it must not be freed before releasing the framebuffer first. If FBs are released first, the crtc using the FB is disabled first so now the GEM object can be freed safely. The CRTC will be enabled again when the driver restores fbdev mode. Signed-off-by: NPrathyush K <prathyush.k@samsung.com> Signed-off-by: NDave Airlie <airlied@redhat.com>
-
由 Jim Meyering 提交于
NUL-terminate after strncpy. If the parameter "profile" has length 16 or more, then strncpy leaves "string" with no NUL terminator, so the following search for '\n' may read beyond the end of that 16-byte buffer. If it finds a newline there, then it will also write beyond the end of that stack buffer. Signed-off-by: NJim Meyering <meyering@redhat.com> Signed-off-by: NDave Airlie <airlied@redhat.com>
-
由 Marcos Paulo de Souza 提交于
The output of "make versioncheck" points a incorrect include of version.h in the drivers/gpu/drm/gma500/mdfld_dsi_output.h: drivers/gpu/drm/gma500/mdfld_dsi_output.h: 32 linux/version.h not needed. If we take a look in the file, we can agree to remove it. Cc: David Airlie <airlied@linux.ie> Cc: <dri-devel@lists.freedesktop.org> Signed-off-by: NMarcos Paulo de Souza <marcos.souza.org@gmail.com> Signed-off-by: NDave Airlie <airlied@redhat.com>
-
由 Jerome Glisse 提交于
If AGP is placed in the middle, the size_af is off-by-one, it results in VRAM being placed at 0x7fffffff instead of 0x8000000. v2: fix the vram_start setup. v3: also fix r7xx & newer ASIC Reported-by: russiane39 on #radeon Signed-off-by: NJerome Glisse <jglisse@redhat.com> Signed-off-by: NDave Airlie <airlied@redhat.com>
-
由 Joonyoung Shim 提交于
Free event and restore event_space only when page_flip->flags has DRM_MODE_PAGE_FLIP_EVENT if page_flip() is failed. Signed-off-by: NJoonyoung Shim <jy0922.shim@samsung.com> Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com> Reviewed-by: NMichel Dänzer <michel.daenzer@amd.com> Signed-off-by: NDave Airlie <airlied@redhat.com>
-
git://people.freedesktop.org/~danvet/drm-intel由 Dave Airlie 提交于
* 'drm-intel-fixes' of git://people.freedesktop.org/~danvet/drm-intel: drm/i915: Do not set "Enable Panel Fitter" on SNB pageflips drm/i915: Hold mode_config lock whilst changing mode for lastclose() drm/i915: don't clobber the special upscaling lvds timings
-
由 Takashi Iwai 提交于
The check of the encoder type in the commit [e00e8b5e: drm/radeon/kms: fix analog load detection on DVI-I connectors] is obviously wrong, and it's the culprit of the regression on my workstation with DVI-analog connection resulting in the blank output. Fixed the typo now. Signed-off-by: NTakashi Iwai <tiwai@suse.de> Reviewed-by: NAlex Deucher <alexander.deucher@amd.com> Cc: <stable@vger.kernel.org> Signed-off-by: NDave Airlie <airlied@redhat.com>
-
由 Dave Airlie 提交于
Allows this module to load correctly with certain debugging options on. Reported on irc by scientes Signed-off-by: NDave Airlie <airlied@redhat.com>
-
由 Hugh Dickins 提交于
My 9ce70c02 "memcg: fix deadlock by inverting lrucare nesting" put a nasty little bug into v3.3's version of mem_cgroup_replace_page_cache(), sometimes used for FUSE. Replacing __mem_cgroup_commit_charge_lrucare() by __mem_cgroup_commit_charge(), I used the "pc" pointer set up earlier: but it's for oldpage, and needs now to be for newpage. Once oldpage was freed, its PageCgroupUsed bit (cleared above but set again here) caused "Bad page state" messages - and perhaps worse, being missed from newpage. (I didn't find this by using FUSE, but in reusing the function for tmpfs.) Signed-off-by: NHugh Dickins <hughd@google.com> Cc: stable@vger.kernel.org [v3.3 only] Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
-