On Sat, Aug 03, 2024 at 01:21:52AM +0000, Wei Liu wrote:

On Fri, Jul 26, 2024 at 03:59:06PM -0700, Roman Kisel wrote:

quoted

To run in the VTL mode, Hyper-V drivers have to know what
VTL the system boots in, and the arm64/hyperv code does not
have the means to compute that.

Refactor the code to hoist the function that detects VTL,
make it arch-neutral to be able to employ it to get the VTL
on arm64. Fix the hypercall output address in `get_vtl(void)`
not to overlap with the hypercall input area to adhere to
the Hyper-V TLFS.

Can you split the fix out? That potentially can be backported.

quoted

Signed-off-by: Roman Kisel <redacted>
---
+
+#if IS_ENABLED(CONFIG_HYPERV_VTL_MODE)
+u8 __init get_vtl(void)
+{
+	u64 control = HV_HYPERCALL_REP_COMP_1 | HVCALL_GET_VP_REGISTERS;
+	struct hv_get_vp_registers_input *input;
+	struct hv_get_vp_registers_output *output;
+	unsigned long flags;
+	u64 ret;
+
+	local_irq_save(flags);
+	input = *this_cpu_ptr(hyperv_pcpu_input_arg);
+	output = *this_cpu_ptr(hyperv_pcpu_output_arg);

Hmm... I don't remember why the old code didn't use
hyperv_pcpu_output_arg but instead reused input+OFFSET as output.

Saurabh, can you comment on this?

This was done to optimize memory usage. Michael has provided more
details on this in his review of the patch today.