Improve the svm fifo allocator

- Round up requested fifo size to the next power of two
- Maintain per-segment power-of-two freelists
- Allocate fifos in chunks, to amortize alignment overhead
- Detach builtin test client application after each run
  so we can use different fifo sizes each time
- Be more suspicious of session / application indices

Useful prep work for dynamically resizing fifos. As far as the svm
fifo code is concerned, it's OK to set fifo->nitems anywhere in the
interval: [0, 1<<(fifo->freelist_index) + FIFO_SEGMENT_MIN_FIFO_SIZE]

It's unlikely that setting nitems below the path MTU will work out
very well...

Change-Id: Idad73a027dfb7412056cb02988b77e300fa7e8a7
Signed-off-by: Dave Barach <dave@barachs.net>

1 files changed, 4 insertions, 1 deletions

diff --git a/src/svm/svm_fifo.c b/src/svm/svm_fifo.c
index e478c06e891..7f8127cfa69 100644
--- a/src/svm/svm_fifo.c
+++ b/src/svm/svm_fifo.c
@@ -192,8 +192,11 @@ svm_fifo_t *
 svm_fifo_create (u32 data_size_in_bytes)
 {
   svm_fifo_t *f;
+  u32 rounded_data_size;
 
-  f = clib_mem_alloc_aligned_or_null (sizeof (*f) + data_size_in_bytes,
+  /* always round fifo data size to the next highest power-of-two */
+  rounded_data_size = (1 << (max_log2 (data_size_in_bytes)));
+  f = clib_mem_alloc_aligned_or_null (sizeof (*f) + rounded_data_size,
 				      CLIB_CACHE_LINE_BYTES);
   if (f == 0)
     return 0;