cpu: x64: brgemm bwd_w convolution: fix batchsizes indexes

Author: ankalinin

src/cpu/x64/jit_brgemm_conv_bwd_w.cpp

@@ -98,7 +98,8 @@ status_t brgemm_convolution_bwd_weights_t::pd_t::init(engine_t *engine) {
         auto M = (i) ? jcp_.M_tail : jcp_.M;
         if (M <= 0) continue;
         // init only needed brgemm descriptors
-        for (int bs = 0; bs <= jcp_.max_batch; bs++) {
+        const auto bs_end = jcp_.var_bs ? 1 : jcp_.max_batch;
+        for (int bs = 0; bs <= bs_end; bs++) {
             if (batchsizes[bs] == -1) continue;
             for_(int i_init = init_begin; i_init < init_end; i_init++)
             for_(int i_N = N_begin; i_N < N_end; i_N++)
@@ -274,7 +275,8 @@ status_t brgemm_convolution_bwd_weights_t::init(engine_t *engine) {
     int init_begin = 0;
     int init_end = 2;
 
-    for (int bs = 0; bs <= jcp.max_batch; bs++) {
+    const auto bs_end = jcp.var_bs ? 1 : jcp.max_batch;
+    for (int bs = 0; bs <= bs_end; bs++) {
         if (_pd->batchsizes[bs] == -1) continue;
 
         for_(int i_N = N_begin; i_N < N_end; i_N++)