Reimplement the read and the write operations

[HotMercury]

problem:
Currently, simplefs does not implement read and write operations and therefore uses the default implementation provided by the Linux kernel. Unfortunately, this implementation assumes that the data blocks are filled contiguously, which leads to potential performance and usability problems.

Implemented simplefs_file_read:

• Reads data from the file based on the specified position and length.
• Ensures that data is read only within the file size limits.
• Directly reads data from the disk through extent blocks.

Implemented simplefs_file_write:

• Writes data to the file starting from the specified position.
• Allocates new blocks if necessary, managing extent-based allocation.
• Updates inode size and block count correctly after writing.
• Ensures data is written within the file system's maximum file size limit.
• Directly writes data to the disk through extent blocks.

Both functions include proper error handling, resource management, and synchronization to ensure data store in disk.

Close #57

[jserv]

Improve the git commit messages. Rewrite in shorter and disciplined manner.

[jserv]

Does the comment "/if exist, read_iter will not be called/" make sense?

[jserv]

Is it time to drop generic_file_read_iter and generic_file_write_iter?

[HotMercury]

I am not sure if it is time to drop these functions. I need to check more. I traced the read syscall and found the following call sequence:
ksys_read -> vfs_read.

if (file->f_op->read)
    ret = file->f_op->read(file, buf, count, pos);
else if (file->f_op->read_iter)
    ret = new_sync_read(file, buf, count, pos);
else
    ret = -EINVAL;

vfs_read

[jserv]

You can utilize Ftrace along with some file system testers such as fio.

[HotMercury]

To test this issue, I used fio and ftrace to check if simplefs and generic_file_read_iter are being used. For convenience, I first wrapped test_generic_file_read_iter. Here are my test results, which show that generic_file_read_iter is indeed not being used.

wrap generic_file_read_iter

static ssize_t test_generic_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
{
    return generic_file_read_iter(iocb, to);
}

fio in simplefs filesystem

$ fio --name=readtest --size=1M --filename=./testfile --bs=4k --rw=read --iodepth=4

ftrace setting

echo function > /sys/kernel/debug/tracing/current_tracer
echo test_generic_file_read_iter > /sys/kernel/debug/tracing/set_ftrace_filter
echo simplefs_read > /sys/kernel/debug/tracing/set_ftrace_filter

We can see that only simplefs_read is used.
output:

root@vm:/sys/kernel/debug/tracing# cat trace
# tracer: function
#
# entries-in-buffer/entries-written: 256/256   #P:4
#
#                                _-----=> irqs-off/BH-disabled
#                               / _----=> need-resched
#                              | / _---=> hardirq/softirq
#                              || / _--=> preempt-depth
#                              ||| / _-=> migrate-disable
#                              |||| /     delay
#           TASK-PID     CPU#  |||||  TIMESTAMP  FUNCTION
#              | |         |   |||||     |         |
             fio-44553   [001] ...1. 83186.031300: simplefs_read <-vfs_read
             fio-44553   [001] ...1. 83186.031325: simplefs_read <-vfs_read
             fio-44553   [001] ...1. 83186.031328: simplefs_read <-vfs_read
             fio-44553   [001] ...1. 83186.031332: simplefs_read <-vfs_read
             ...
             fio-44553   [001] ...1. 83186.031880: simplefs_read <-vfs_read

[jserv]

We can see that only simplefs_read is used.

It sounds good. Let's move forward.

[HotMercury]

Should I drop them in this PR?

[jserv]

Should I drop them in this PR?

Yes. And you can append the above experiments and observations.

[jserv]

Thank @HotMercury for contributing!