2022-09-30 05:41:04 +02:00
|
|
|
/****************************************************************************
|
|
|
|
* apps/system/cachespeed/cachespeed_main.c
|
|
|
|
*
|
|
|
|
* Licensed to the Apache Software Foundation (ASF) under one or more
|
|
|
|
* contributor license agreements. See the NOTICE file distributed with
|
|
|
|
* this work for additional information regarding copyright ownership. The
|
|
|
|
* ASF licenses this file to you under the Apache License, Version 2.0 (the
|
|
|
|
* "License"); you may not use this file except in compliance with the
|
|
|
|
* License. You may obtain a copy of the License at
|
|
|
|
*
|
|
|
|
* http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
*
|
|
|
|
* Unless required by applicable law or agreed to in writing, software
|
|
|
|
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
|
|
|
|
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
|
|
|
|
* License for the specific language governing permissions and limitations
|
|
|
|
* under the License.
|
|
|
|
*
|
|
|
|
****************************************************************************/
|
|
|
|
|
|
|
|
/****************************************************************************
|
|
|
|
* Included Files
|
|
|
|
****************************************************************************/
|
|
|
|
|
|
|
|
#include <nuttx/arch.h>
|
|
|
|
#include <nuttx/cache.h>
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
#include <nuttx/config.h>
|
2022-09-30 05:41:04 +02:00
|
|
|
#include <nuttx/irq.h>
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
#include <inttypes.h>
|
|
|
|
#include <malloc.h>
|
2022-09-30 05:41:04 +02:00
|
|
|
#include <stdio.h>
|
|
|
|
#include <stdlib.h>
|
|
|
|
#include <time.h>
|
|
|
|
|
|
|
|
/****************************************************************************
|
|
|
|
* Pre-processor Definitions
|
|
|
|
****************************************************************************/
|
|
|
|
|
|
|
|
#define CACHESPEED_PREFIX "CACHE Speed: "
|
2023-06-01 14:12:22 +02:00
|
|
|
#define REPEAT_NUM 1000
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
#ifdef CACHESPEED_PERFTIME
|
|
|
|
#define TIME uint64_t
|
|
|
|
|
|
|
|
#define CONVERT(cost) \
|
|
|
|
do \
|
|
|
|
{ \
|
|
|
|
struct timespec ts; \
|
2023-10-06 08:01:08 +02:00
|
|
|
perf_convert(cost, &ts); \
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
cost = ts.tv_sec * 1000000000 + ts.tv_nsec; \
|
|
|
|
} while (0)
|
|
|
|
|
2023-10-06 08:01:08 +02:00
|
|
|
#define TIMESTAMP(x) (x) = perf_gettime()
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
#else
|
|
|
|
#define TIME time_t
|
|
|
|
|
|
|
|
#define CONVERT(cost)
|
|
|
|
|
|
|
|
#define TIMESTAMP(x) \
|
2022-09-30 05:41:04 +02:00
|
|
|
do \
|
|
|
|
{ \
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
struct timespec ts; \
|
|
|
|
clock_gettime(CLOCK_MONOTONIC, &ts); \
|
|
|
|
x = ts.tv_sec * 1000000000 + ts.tv_nsec; \
|
2022-09-30 05:41:04 +02:00
|
|
|
} while (0)
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
#endif
|
|
|
|
|
|
|
|
#define GET_DCACHE_LINE up_get_dcache_linesize()
|
|
|
|
#define GET_ICACHE_LINE up_get_icache_linesize()
|
|
|
|
#define GET_DCACHE_SIZE up_get_dcache_size()
|
|
|
|
#define GET_ICACHE_SIZE up_get_icache_size()
|
2022-09-30 05:41:04 +02:00
|
|
|
|
|
|
|
/****************************************************************************
|
|
|
|
* Private Types
|
|
|
|
****************************************************************************/
|
|
|
|
|
|
|
|
struct cachespeed_s
|
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
uintptr_t addr;
|
|
|
|
size_t alloc;
|
2022-09-30 05:41:04 +02:00
|
|
|
};
|
|
|
|
|
|
|
|
/****************************************************************************
|
|
|
|
* Private Data
|
|
|
|
****************************************************************************/
|
|
|
|
|
|
|
|
/****************************************************************************
|
|
|
|
* Private Functions
|
|
|
|
****************************************************************************/
|
|
|
|
|
|
|
|
/****************************************************************************
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
* Name: setup
|
2022-09-30 05:41:04 +02:00
|
|
|
****************************************************************************/
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
static void setup(FAR struct cachespeed_s *cs)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
struct mallinfo info = mallinfo();
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
/* Get the currently available memory from the system. We want the
|
|
|
|
* memset range to be as large as possible in our tests to ensure
|
|
|
|
* that the cache is filled with our dirty data
|
|
|
|
*/
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
cs->alloc = info.fordblks / 2;
|
|
|
|
cs->addr = (uintptr_t)malloc(cs->alloc);
|
|
|
|
if (cs->addr == 0)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
printf(CACHESPEED_PREFIX "Unable to request memory.\n");
|
|
|
|
exit(EXIT_FAILURE);
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
/* Let's export the test message */
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
printf(CACHESPEED_PREFIX "address src: %" PRIxPTR "\n", cs->addr);
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/****************************************************************************
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
* Name: teardown
|
2022-09-30 05:41:04 +02:00
|
|
|
****************************************************************************/
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
static void teardown(FAR struct cachespeed_s *cs)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
free((void *)cs->addr);
|
|
|
|
printf(CACHESPEED_PREFIX "Done!\n");
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/****************************************************************************
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
* Name: report_line
|
2022-09-30 05:41:04 +02:00
|
|
|
****************************************************************************/
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
static void report_line(size_t bytes, TIME cost)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
double rate;
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
/* There is a situation: if the time is 0, then the
|
|
|
|
* calculated speed is wrong.
|
|
|
|
*/
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
CONVERT(cost);
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
if (cost == 0)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
printf(CACHESPEED_PREFIX "%d bytes cost time too small!\n", bytes);
|
|
|
|
return;
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
/* rate = Test Data Size / Execution Time */
|
|
|
|
|
|
|
|
rate = 1.00 * bytes * REPEAT_NUM / cost;
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
printf("%d Bytes: %4lf, %4llu, %4llu\n\r",
|
|
|
|
bytes, rate, cost / REPEAT_NUM, cost);
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/****************************************************************************
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
* Name: test_skeleton
|
2022-09-30 05:41:04 +02:00
|
|
|
****************************************************************************/
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
static void test_skeleton(FAR struct cachespeed_s *cs,
|
|
|
|
const size_t cache_size,
|
|
|
|
const size_t cache_line_size, int align,
|
|
|
|
void (*func)(uintptr_t, uintptr_t),
|
|
|
|
const char *name)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
size_t update_size;
|
|
|
|
printf("** %s [rate, avg, cost] in nanoseconds(bytes/nesc) %s **\n",
|
|
|
|
name, align ? "align" : "unalign");
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
if (!align)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
update_size = cache_line_size - 1;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
update_size = cache_line_size;
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
|
2023-06-01 14:12:22 +02:00
|
|
|
for (size_t bytes = update_size;
|
|
|
|
bytes <= cache_size; bytes = 2 * bytes)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
irqstate_t irq;
|
|
|
|
TIME start;
|
|
|
|
TIME end;
|
|
|
|
TIME cost = 0;
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
/* Make sure that test with all the contents
|
|
|
|
* of our address in the cache.
|
|
|
|
*/
|
|
|
|
|
|
|
|
up_flush_dcache_all();
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
irq = enter_critical_section();
|
|
|
|
for (int i = 0; i < REPEAT_NUM; i++)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
memset((void *)cs->addr, 1, cs->alloc);
|
|
|
|
TIMESTAMP(start);
|
|
|
|
func(cs->addr, (uintptr_t)(cs->addr + bytes));
|
|
|
|
TIMESTAMP(end);
|
|
|
|
cost += end - start;
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
leave_critical_section(irq);
|
|
|
|
report_line(bytes, cost);
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/****************************************************************************
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
* Name: cachespeed_common
|
2022-09-30 05:41:04 +02:00
|
|
|
****************************************************************************/
|
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
static void cachespeed_common(struct cachespeed_s *cs)
|
2022-09-30 05:41:04 +02:00
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
test_skeleton(cs, GET_DCACHE_SIZE, GET_DCACHE_LINE, 1,
|
|
|
|
up_invalidate_dcache, "dcache invalidate");
|
|
|
|
test_skeleton(cs, GET_DCACHE_SIZE, GET_DCACHE_LINE, 0,
|
|
|
|
up_invalidate_dcache, "dcache invalidate");
|
|
|
|
test_skeleton(cs, GET_DCACHE_SIZE, GET_DCACHE_LINE, 1,
|
|
|
|
up_clean_dcache, "dcache clean");
|
|
|
|
test_skeleton(cs, GET_DCACHE_SIZE, GET_DCACHE_LINE, 0,
|
|
|
|
up_clean_dcache, "dcache clean");
|
|
|
|
test_skeleton(cs, GET_DCACHE_SIZE, GET_DCACHE_LINE, 1,
|
|
|
|
up_flush_dcache, "dcache flush");
|
|
|
|
test_skeleton(cs, GET_DCACHE_SIZE, GET_DCACHE_LINE, 0,
|
|
|
|
up_flush_dcache, "dcache flush");
|
|
|
|
test_skeleton(cs, GET_ICACHE_SIZE, GET_ICACHE_LINE, 1,
|
|
|
|
up_invalidate_icache, "icache invalidate");
|
|
|
|
test_skeleton(cs, GET_ICACHE_SIZE, GET_ICACHE_LINE, 0,
|
|
|
|
up_invalidate_icache, "icache invalidate");
|
2022-09-30 05:41:04 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/****************************************************************************
|
|
|
|
* Public Functions
|
|
|
|
****************************************************************************/
|
|
|
|
|
|
|
|
/****************************************************************************
|
|
|
|
* Name: cachespeed_main
|
|
|
|
****************************************************************************/
|
|
|
|
|
|
|
|
int main(int argc, FAR char *argv[])
|
|
|
|
{
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
struct cachespeed_s cs =
|
|
|
|
{
|
|
|
|
.addr = 0,
|
|
|
|
.alloc = 0
|
|
|
|
};
|
2022-09-30 05:41:04 +02:00
|
|
|
|
system/cachespeed:Redesign of the overall test tool.
Compared to the previous version, the following factors have been taken into account in the new version of cachespeed:
1. the user needs to care about getting a piece of available memory (even if this piece of memory may memory stomp during use)
2. High repetition of test code, poor readability, and overall high test coupling
3. Not taking into account the alignment cache line situation
4. When comparing values between different test tools, there is a little error because of the different ways to get the values (different fine reading)
Therefore, in the new version according to the above problems have been improved:
1. Provide a "test_skeleton", which contains the overall testing process
2. The user only needs to execute, not to care about how to get the available memory address, and the memory allocated by the program also ensures the memory security and data accuracy
3. The system to obtain the data required for the test reduces the difficulty of use and the possibility of inaccurate results due to data errors.
4. Provide two kinds of precision data results, which can be configured through Kconfig
5. Optimize the output log, now more intuitive and concise, to help the subsequent data organization and observation
6. New test items for aligned/unaligned cache line
7. Better readability and extensibility, making it easier to add/remove test items
Signed-off-by: chenrun1 <chenrun1@xiaomi.com>
2023-05-05 12:11:16 +02:00
|
|
|
setup(&cs);
|
|
|
|
cachespeed_common(&cs);
|
|
|
|
teardown(&cs);
|
2022-09-30 05:41:04 +02:00
|
|
|
return 0;
|
|
|
|
}
|