Made by Timothy Manning <timothy@yaffs.net> on 04/11/2010
-the yaffs_open function has been covered by tests
-the yaffs_close function has been covered by tests
-
-tests made
+Tests made
test_yaffs_mount
+ test_yaffs_mount_ENODEV
+ test_yaffs_mount_ENAMETOOLONG
+ test_yaffs_mount_ENOENT
+ test_yaffs_mount_EBUSY //called when trying to mount a new mount point with a mount point already mounted.
+
test_yaffs_access
- test_yaffs_close
+ test_yaffs_access_ENIVAL //when the mode is incorrect.
+ test_yaffs_access_ENOTDIR
+ test_yaffs_access_ENOENT
+
test_yaffs_close_EBADF
+
test_yaffs_ftruncate
+ test_yaffs_ftruncate_EBADF
+ test_yaffs_ftruncate_ENIVAL
+ test_yaffs_ftruncate_EFBIG
+
test_yaffs_lseek
+ test_yaffs_lseek_EBADF
+ test_yaffs_lseek_EINVAL
+ test_yaffs_lseek_EFBIG
+
test_yaffs_open
test_yaffs_open_EEXIST
test_yaffs_open_EISDIR
test_yaffs_open_ENAMETOOLONG
test_yaffs_open_ENOENT
test_yaffs_open_ENOTDIR
+ test_yaffs_open_EINVAL
+ test_yaffs_open_EINVAL2 //the function open has two modes which can be set, so two tests modes are needed.
+
test_yaffs_read
+ test_yaffs_read_EBADF
+
test_yaffs_stat
+
test_yaffs_truncate
+ test_yaffs_truncate_ENOTDIR
+ test_yaffs_truncate_EISDIR
+ test_yaffs_truncate_ENOENT
+ test_yaffs_truncate_ENIVAL
+ test_yaffs_truncate_EFBIG
+
+ test_yaffs_unlink
test_yaffs_unlink_EISDIR
test_yaffs_unlink_ENAMETOOLONG
test_yaffs_unlink_ENOENT
test_yaffs_unlink_ENOTDIR
+ test_yaffs_unlink_ENOENT
+
+ test_yaffs_unmount
test_yaffs_write
-tests to add
- test_yaffs_mount_EACCES
- test_yaffs_mount_EINVAL
- test_yaffs_mount_ELOOP
- test_yaffs_mount_EMFILE
- test_yaffs_mount_ENAMETOOLONG
- test_yaffs_mount_ENOENT
- test_yaffs_mount_ENOTDIR
+
+ add a test where a directory is moved. /fluffy/box. move "fluffy" to "/fluffy/frog".
+
+Tests to add
+ test_yaffs_mount_EACCES //Cannot be generated with yaffs.
+ test_yaffs_mount_EINVAL //Cannot be generated with yaffs.
+ test_yaffs_mount_ELOOP //Cannot be generated with yaffs.
+ test_yaffs_mount_EMFILE //Cannot be generated with yaffs.
+ test_yaffs_mount_ENOTDIR //Cannot be generated with yaffs.
+
+ test_yaffs_umount_ENODEV
+ test_yaffs_umount_ENAMETOOLONG
+ test_yaffs_umount_ENOENT
+ test_yaffs_umount_EBUSY
+
test_yaffs_open_EACCES
test_yaffs_open_ENOSPC
- test_yaffs_open_ELOOP Too many symbolic links were encountered in resolving pathname
- test_yaffs_close
+ test_yaffs_open_ELOOP //Too many symbolic links were encountered in resolving pathname
test yaffs_open_running_out_of_handles error
+
+ test_yaffs_close //This function has already been called by the time this test is reached.
+
+
+ test_yaffs_read_EINVAL //test exists but is not finshed yet.
+ test_yaffs_read_EISDIR //Cannot be generated with yaffs.
+ test what happens if you read off the end of the file?
+
test_yaffs_unlink_EACCES
test_yaffs_unlink_ELOOP
- test_yaffs_unlink_ENOENT
test_yaffs_unlink_ENOMEM
- check to see if an error code is generated when there isn't an error
+
test_yaffs_access_EACCESS
test_yaffs_access_ELOOP
test_yaffs_access_ENAMETOOLONG
- test_yaffs_access_ENOENT
- test_yaffs_access_ENOTDIR
- test_yaffs_access_ENIVAL //mode is incorrect.
+ test_yaffs_access_ENOENT_generated_with_a_dangling_symbloic_link
+
+ test_yaffs_ftruncate_EACCES
+ test_yaffs_ftruncate_EISDIR //Cannot be generated with yaffs.
+ test_yaffs_ftruncate_ELOOP //Cannot be generated with yaffs.
+ test_yaffs_ftruncate_ENOENT //Cannot be generated with yaffs.
+ test_yaffs_ftruncate_ENOTDIR //Cannot be generated with yaffs.
+
+ test_yaffs_truncate_EACCES
+ test_yaffs_truncate_ELOOP
+ test_yaffs_truncate_ENAMETOOLONG
+ add a truncate function for truncating a file size to -1.
+
+ What happens if a handle is opened to a file and the file is then deleted?
+ Check to see if yaffs generates an error code for no reason.
+ What happens when a file is opened with no modes set?
How to add a test
Test_yaffs_[function of yaffs which is been tested]_[error trying to be generated]
The .c file needs to contain two functions.
- The first function needs to contain the code for the main test and will return -1 on a failure and 0 or greater on a success.
- The second function needs contain the code for cleaning up after the test. Cleaning up may include closing some open handles, recreating a file, ect.
+ The first function needs to contain the code for the main test and will
+ return -1 on a failure and 0 or greater on a success.
+ The second function needs contain the code for cleaning up after the test.
+ Cleaning up may include closing some open handles, recreating a file, ect.
This second function needs to return -1 on a failure and 0 or greater on success.
- The name of first function needs to be called the same as the file name (without the .c or .h)
- The second function's name needs be the same as the first function but with "_clean" added on the end.
+ The name of first function needs to be called the same as the file
+ name (without the .c or .h)
+ The second function's name needs be the same as the first function but
+ with "_clean" added on the end.
- So if a test is been created for the yaffs function yaffs_fish() then create these files
- Test_yaffs_fish.c
- Contains int test_yaffs_fish(void); int test_yaffs_fish_clean(void);
- Test_yaffs_fish.h
+ So if a test is been created for the yaffs function yaffs_foo() then
+ create these files
+ test_yaffs_foo.c
+ Contains int test_yaffs_foo(void); int test_yaffs_foo_clean(void);
+ test_yaffs_foo.h
Which includes "lib.h", "yaffsfs.h" header files.
Next write the test code in these files then add these files to the Makefile.
- Add the name of the test files' object file (test_yaffs_fish.o ) to the TESTFILES tag around line 50 of the Makefile.
-
-
+ Add the name of the test files' object file (test_yaffs_foo.o ) to the
+ TESTFILES tag around line 50 of the Makefile.
Now add the test functions to the test_list[] array in quick_tests.h
- The order of the tests matters. The idea is to test each yaffs_function individualy and only using tested yaffs_components before using this new yaffs_function.
- This array consists of: {[test function], [the clean function], [name of the tests which will be printed when the test fails]},
+ The order of the tests matters. The idea is to test each yaffs_function
+ individualy and only using tested yaffs_components before using this new
+ yaffs_function.
+ This array consists of:
+ {[test function], [the clean function], [name of the tests which will be printed when the test fails]},
- So add this line to the test_list[]: {test_yaffs_fish, test_yaffs_fish_clean, "test_yaffs_fish"},
+ So add this line to the test_list[]: {test_yaffs_foo, test_yaffs_foo_clean, "test_yaffs_foo"},
- Also include the test's .h file in the quick_test.h file: #include "test_yaffs_fish.h"
-
- The test file should now make and run.
-
-
+ Also include the test's .h file in the quick_test.h file: #include "test_yaffs_foo.h"
+ The test file should now make and run(you may need to make clean first).
- PS: yaffs_fish() is a made up function for this README (in case you want to try and find this function in yaffs).
+ PS: yaffs_foo() is a made up function for this README (in case you want
+ to find this function in yaffs).
-BUGS
- Needing to include int random_seed; and int simulate_power_failure = 0; in any main program using yaffsfs.h
- bug with opening a file with a name of 1,000,000 char long with no errors.
- bug with unlinking a file with 1,000,000 get the error ENOENT but should be geting ENAMETOOLONG.
-
- FIXED-ENOSPC error in programs test_yaffs_open_ENOTDIR and test_yaffs_open_ENOENT.
- FIXED-ENOENT been returned by yaffs_read but the handle is good and the yaffs_open function does not return an error.
projects / yaffs2.git / blobdiff