According to SDM, we should check whether the gva access from guest is valid. If it's not, correct exception should be injected. We only need to emulate the instructions which access the memory and

I tried android UOS and there is no VM exit with lock prefix. And I suppose we could drop this patch because lock prefix only could be used with few instructions. It's very unlikely MMIO operation has

Acked-by: Eddie Dong <eddie.dong@...>

Please move forward... Acked-by: Eddie Dong <eddie.dong@...>

Have updated and PRed. Thanks. -Yonghua

Acked-by: Eddie Dong <eddie.dong@...> BTW, it is not necessary to initialize it to 0.

By looking at the data structure: struct irq_desc { uint32_t irq; /* index to irq_desc_base */ enum irq_state used; /* this irq have assigned to device */ enum

Hi Eddie, Eddie Dong <eddie.dong@...> writes:

Reviewed. Patch is good to me.

Reviewed-by: Edwin Zhai <edwin.zhai@...> Best Rgds, Edwin

The guc boot option is refined on the new linux kernel. The boot option of "i915.enable_guc=0" should be added in order to disable Guc instead of using "enable_guc_loading/submission". But in order to

THis is the patch set that adds one boot option to disable Guc on SOS/UOS. The guc boot option is refined on the new linux kernel. The boot option of "i915.enable_guc=0" should be added in order to

The guc boot option is refined on the new linux kernel. The boot option of "i915.enable_guc=0" should be added in order to disable Guc instead of using "enable_guc_loading/submission". But in order to

Reviewed-by: Minggui Cao <minggui.cao@...>

This patch adds code to support read-only RTC support for guests run by partition mode ACRN. It supports RW for CMOS address port 0x70 and RO for CMOS data port 0x71. Reads to CMOS RAM offsets are

- remove the hardcoding loccal variable name 'cpu_int_vale' Signed-off-by: Yonghua Huang <yonghua.huang@...> --- hypervisor/arch/x86/ioapic.c | 13 ++++--- hypervisor/arch/x86/irq.c

Shiqing Gao <shiqing.gao@...> writes: Reviewed-by: Junjie Mao <junjie.mao@...> For the rest violations related to integral types, please follow up with another patch. Thanks. -- Best

It looks quite similar with acrn log, I am think if we could removed one, or merge them. shall we have an offline discussion first? Thanks, Like