[siemens/jailhouse] aa5fd1: driver: Make a bunch of cell functions static

2016-07-31 Thread GitHub
Branch: refs/heads/next Home: https://github.com/siemens/jailhouse Commit: aa5fd101ec96690168810d4e0c0c4756cd2ce3a8 https://github.com/siemens/jailhouse/commit/aa5fd101ec96690168810d4e0c0c4756cd2ce3a8 Author: Jan Kiszka Date: 2016-07-31 (Sun, 31 Jul

Re: [PATCH 12/12] x86: Require VT-d support also for QEMU

2016-07-31 Thread Jan Kiszka
On 2016-07-31 10:39, Jan Kiszka wrote: > From: Jan Kiszka > > QEMU 2.7 will have all the features we need to run Jailhouse with VT-d > IOMMU. Remove the special cases from the code and document the new > command line. > > The requirement on the kernel version is not

Re: Passing parameters into inmate through the cmdline

2016-07-31 Thread Jan Kiszka
On 2016-07-30 16:38, Maxim Baryshnikov wrote: > Hello everyone, > > I am trying to use the cmdline feature in Jailhouse, but it seems > that it does not work :( For example, when I load apic-demo like this: > > #jailhouse cell load apic-demo -a 0xf -s > "pollute_cache" > > Then,

[PATCH 11/12] configs: Adjust QEMU config to upcoming release 2.7

2016-07-31 Thread Jan Kiszka
From: Jan Kiszka While 2.7 is still in -rc, we will significantly benefit from this release: it comes with VT-d support that now includes interrupt remapping. Therefore move forward to this QEMU version. Signed-off-by: Jan Kiszka --- README.md

[PATCH 01/12] configs: Expand ivshmem regions to 1 MB

2016-07-31 Thread Jan Kiszka
From: Jan Kiszka More space helps with throughput when using the shared memory for inter-cell networking. Signed-off-by: Jan Kiszka --- configs/ivshmem-demo.c | 6 +++--- configs/linux-x86-demo.c | 8 configs/qemu-vm.c| 8

[PATCH 04/12] core: Deny shutdown with still existing non-root cells

2016-07-31 Thread Jan Kiszka
From: Jan Kiszka Simplify the shutdown logic by requiring the root cell driver to explicitly destroy all non-root cells first. In particular, it helps to resolve the tricky handover of non-root cell CPUs in unplugged state to Linux. At this chance, properly document

[PATCH 07/12] driver: Gracefully handle zero-sized load requests

2016-07-31 Thread Jan Kiszka
From: Jan Kiszka We currently fail if a load request has a size of 0, because mapping an empty region fails. While such requests are often caused by corrupted files there may also be valid use cases. So better handle this gracefully. Signed-off-by: Jan Kiszka

[PATCH 00/12] Assorted fixes and simplifications

2016-07-31 Thread Jan Kiszka
Besides config updates and some cleanups, this series comes with two major changes. The most user visible one is demanding VT-d support. QEMU 2.7 is currently in -rc and will deliver all features we need to run Jailhouse. Therefore, we can drop some special cases from the VT-d code by demanding

[PATCH 12/12] x86: Require VT-d support also for QEMU

2016-07-31 Thread Jan Kiszka
From: Jan Kiszka QEMU 2.7 will have all the features we need to run Jailhouse with VT-d IOMMU. Remove the special cases from the code and document the new command line. The requirement on the kernel version is not exact but a safe guess based on tests. Support for

[PATCH 06/12] core: Fix printing of invalid MMIO accesses

2016-07-31 Thread Jan Kiszka
From: Jan Kiszka mem->phys_start is u64, but we only print up to unsigned long. Adjust the input type. Resolves broken size reports on 32-bit ARM. Signed-off-by: Jan Kiszka --- hypervisor/mmio.c | 3 ++- 1 file changed, 2 insertions(+), 1

[PATCH 08/12] arm: Remove return code from irqchip_cpu_reset

2016-07-31 Thread Jan Kiszka
From: Jan Kiszka Only the GICv3 actually returns a non-zero value, and that only if the reset was called before the irqchip was fully initialized, i.e. as part of an error rollback triggered via irqchip_cpu_shutdown. Therefore, remove the error return code and only avoid