mshv: Redesign hypervisor memory deposit logic

The current memory deposit implementation has a few issues and bugs:
 o It is very slow: one of the main contributions to slow VM creations
   and boot is many hypercalls repeatedly coming back with insufficient
   memory. On Hyper-V, a hypercall returns with such status whenever it
   cannot complete due to lack of memory in the hypervisor and needs more.
 o Contiguous range requirement is broken, and is critical bug
 o An incorrect assumption is made that contiguous memory size would
   always be power of 2.
 o Two pages are allocated, only one is really needed. Extra alloc adds
   to the overhead.
 o For a 512 page deposit, the allocation is split into two: one for 511
   and second for 1. Thus, an order 9 allocation never happens. A
   contiguous 2M range, if possible, greatly improves performance
   in the hypervisor as it can be mapped as large page whenever possible.
   Also, removing extra allocation reduces overhead in linux.
 o Since a page is already allocated to collect the frames, there is
   not really a need to use per cpu input page, and hence avoid local
   irq disable.
 o In hv_call_deposit_pages(), in case of error, under err_free_allocations
   label, all pages are freed without checking status to see if some pages
   were deposited. This is a critical bug as it would free pages that hyp
   may be using.

All of above is addressed by:
 o Allocate 2M by default, this is the recommendation from the hypervisor
   team, and greatly improves performance. Depositing one or few pages
   at a time results in lot of insufficient memory returns from hypercalls.
 o Always start with a full 2M range allocation, thus getting contiguous if
   available. In cases where possible, the deposits are much faster.
 o Allocate only one page in the deposit function and collect 511 pfns
   there. Just use a local variable for last pfn.
 o Use the page as input to hypercall. Since this page is locally allocated,
   irq disable can be avoided helping speed up the deposit.
 o Fix the physical contiguous memory enforcement..
 o Lastly, remove pre-deposits hv_call_create_vp() and
   hv_call_initialize_partition() as they were removed internally while
   ago, most likely because they didn't help much.

Signed-off-by: Mukesh R <mrathor@linux.microsoft.com>
Reviewed-by: Souradeep Chakrabarti <schakrabarti@linux.microsoft.com>
Signed-off-by: Wei Liu <wei.liu@kernel.org>

1 files changed, 0 insertions, 5 deletions

diff --git a/include/asm-generic/mshyperv.h b/include/asm-generic/mshyperv.h
index 98fe5c0f54ea..064c40bb2821 100644
--- a/include/asm-generic/mshyperv.h
+++ b/include/asm-generic/mshyperv.h
@@ -343,7 +343,6 @@ static inline bool hv_parent_partition(void)
 
 bool hv_result_needs_memory(u64 status);
 int hv_deposit_memory_node(int node, u64 partition_id, u64 status);
-int hv_call_deposit_pages(int node, u64 partition_id, u32 num_pages);
 int hv_call_add_logical_proc(int node, u32 lp_index, u32 acpi_id);
 int hv_call_notify_all_processors_started(void);
 bool hv_lp_exists(u32 lp_index);
@@ -358,10 +357,6 @@ static inline int hv_deposit_memory_node(int node, u64 partition_id, u64 status)
 {
 	return -EOPNOTSUPP;
 }
-static inline int hv_call_deposit_pages(int node, u64 partition_id, u32 num_pages)
-{
-	return -EOPNOTSUPP;
-}
 static inline int hv_call_add_logical_proc(int node, u32 lp_index, u32 acpi_id)
 {
 	return -EOPNOTSUPP;