|
4157 | 4157 | rcu_node tree with an eye towards determining
|
4158 | 4158 | why a new grace period has not yet started.
|
4159 | 4159 |
|
4160 |
| - rcuperf.gp_async= [KNL] |
| 4160 | + rcuscale.gp_async= [KNL] |
4161 | 4161 | Measure performance of asynchronous
|
4162 | 4162 | grace-period primitives such as call_rcu().
|
4163 | 4163 |
|
4164 |
| - rcuperf.gp_async_max= [KNL] |
| 4164 | + rcuscale.gp_async_max= [KNL] |
4165 | 4165 | Specify the maximum number of outstanding
|
4166 | 4166 | callbacks per writer thread. When a writer
|
4167 | 4167 | thread exceeds this limit, it invokes the
|
4168 | 4168 | corresponding flavor of rcu_barrier() to allow
|
4169 | 4169 | previously posted callbacks to drain.
|
4170 | 4170 |
|
4171 |
| - rcuperf.gp_exp= [KNL] |
| 4171 | + rcuscale.gp_exp= [KNL] |
4172 | 4172 | Measure performance of expedited synchronous
|
4173 | 4173 | grace-period primitives.
|
4174 | 4174 |
|
4175 |
| - rcuperf.holdoff= [KNL] |
| 4175 | + rcuscale.holdoff= [KNL] |
4176 | 4176 | Set test-start holdoff period. The purpose of
|
4177 | 4177 | this parameter is to delay the start of the
|
4178 | 4178 | test until boot completes in order to avoid
|
4179 | 4179 | interference.
|
4180 | 4180 |
|
4181 |
| - rcuperf.kfree_rcu_test= [KNL] |
| 4181 | + rcuscale.kfree_rcu_test= [KNL] |
4182 | 4182 | Set to measure performance of kfree_rcu() flooding.
|
4183 | 4183 |
|
4184 |
| - rcuperf.kfree_nthreads= [KNL] |
| 4184 | + rcuscale.kfree_nthreads= [KNL] |
4185 | 4185 | The number of threads running loops of kfree_rcu().
|
4186 | 4186 |
|
4187 |
| - rcuperf.kfree_alloc_num= [KNL] |
| 4187 | + rcuscale.kfree_alloc_num= [KNL] |
4188 | 4188 | Number of allocations and frees done in an iteration.
|
4189 | 4189 |
|
4190 |
| - rcuperf.kfree_loops= [KNL] |
4191 |
| - Number of loops doing rcuperf.kfree_alloc_num number |
| 4190 | + rcuscale.kfree_loops= [KNL] |
| 4191 | + Number of loops doing rcuscale.kfree_alloc_num number |
4192 | 4192 | of allocations and frees.
|
4193 | 4193 |
|
4194 |
| - rcuperf.nreaders= [KNL] |
| 4194 | + rcuscale.nreaders= [KNL] |
4195 | 4195 | Set number of RCU readers. The value -1 selects
|
4196 | 4196 | N, where N is the number of CPUs. A value
|
4197 | 4197 | "n" less than -1 selects N-n+1, where N is again
|
|
4200 | 4200 | A value of "n" less than or equal to -N selects
|
4201 | 4201 | a single reader.
|
4202 | 4202 |
|
4203 |
| - rcuperf.nwriters= [KNL] |
| 4203 | + rcuscale.nwriters= [KNL] |
4204 | 4204 | Set number of RCU writers. The values operate
|
4205 |
| - the same as for rcuperf.nreaders. |
| 4205 | + the same as for rcuscale.nreaders. |
4206 | 4206 | N, where N is the number of CPUs
|
4207 | 4207 |
|
4208 |
| - rcuperf.perf_type= [KNL] |
| 4208 | + rcuscale.perf_type= [KNL] |
4209 | 4209 | Specify the RCU implementation to test.
|
4210 | 4210 |
|
4211 |
| - rcuperf.shutdown= [KNL] |
| 4211 | + rcuscale.shutdown= [KNL] |
4212 | 4212 | Shut the system down after performance tests
|
4213 | 4213 | complete. This is useful for hands-off automated
|
4214 | 4214 | testing.
|
4215 | 4215 |
|
4216 |
| - rcuperf.verbose= [KNL] |
| 4216 | + rcuscale.verbose= [KNL] |
4217 | 4217 | Enable additional printk() statements.
|
4218 | 4218 |
|
4219 |
| - rcuperf.writer_holdoff= [KNL] |
| 4219 | + rcuscale.writer_holdoff= [KNL] |
4220 | 4220 | Write-side holdoff between grace periods,
|
4221 | 4221 | in microseconds. The default of zero says
|
4222 | 4222 | no holdoff.
|
|
4490 | 4490 | refscale.shutdown= [KNL]
|
4491 | 4491 | Shut down the system at the end of the performance
|
4492 | 4492 | test. This defaults to 1 (shut it down) when
|
4493 |
| - rcuperf is built into the kernel and to 0 (leave |
4494 |
| - it running) when rcuperf is built as a module. |
| 4493 | + refscale is built into the kernel and to 0 (leave |
| 4494 | + it running) when refscale is built as a module. |
4495 | 4495 |
|
4496 | 4496 | refscale.verbose= [KNL]
|
4497 | 4497 | Enable additional printk() statements.
|
|
0 commit comments