fs: define rs flag as O_RDONLY, not O_RDONLY | O_SYNC

[wafuwafu13]

O_SYNC means "Write operations on the file will complete according to the requirements of synchronized I/O file integrity completion"(ref).
So, setting O_RDONLY and O_SYNC at the same time is inconsistent.
According to the following test, rs has the same result as r.

node/deps/uv/test/test-fs-open-flags.c

Lines 276 to 294 in 2a4452a

	/* r */
	flags = add_flags | UV_FS_O_RDONLY;
	openFail(absent_file, UV_ENOENT);
	writeFail(empty_file, UV_EBADF);
	readExpect(empty_file, "", 0);
	writeFail(dummy_file, UV_EBADF);
	readExpect(dummy_file, "a", 1);
	writeFail(empty_dir, UV_EBADF);
	readFail(empty_dir, UV_EISDIR);
	/* rs */
	flags = add_flags | UV_FS_O_RDONLY | UV_FS_O_SYNC;
	openFail(absent_file, UV_ENOENT);
	writeFail(empty_file, UV_EBADF);
	readExpect(empty_file, "", 0);
	writeFail(dummy_file, UV_EBADF);
	readExpect(dummy_file, "a", 1);
	writeFail(empty_dir, UV_EBADF);
	readFail(empty_dir, UV_EISDIR);

Also, rs is not in document https://nodejs.org/api/fs.html#file-system-flags .

[mscdex]

Changes to libuv need to be made upstream, not in this repo.

[mscdex]

It may be better to instead treat 'rs'/'sr' as 'r' for better backwards compatibility, otherwise this could be a breaking change (I'm not sure how widespread 'rs'/'sr' is) since it will start throwing an error.

[aduh95]

/cc @nodejs/fs

[bnoordhuis]

I don't think you can say O_RDONLY and O_RDONLY|O_SYNC are equivalent. O_SYNC implies metadata is written to disk synchronously. That includes the access time.

[wafuwafu13]

I thought it was O_DSYNC that is related to hardware, but now I understand that O_SYNC also has meaning other than reading and writing files.

https://man7.org/linux/man-pages/man2/open.2.html

O_SYNC provides synchronized I/O file integrity completion,
meaning write operations will flush data and all associated
metadata to the underlying hardware. O_DSYNC provides
synchronized I/O data integrity completion, meaning write
operations will flush data to the underlying hardware, but will
only flush metadata updates that are required to allow a
subsequent read operation to complete successfully

Before Linux 2.6.33, Linux implemented only the O_SYNC flag for
open(). However, when that flag was specified, most filesystems
actually provided the equivalent of synchronized I/O data
integrity completion (i.e., O_SYNC was actually implemented as
the equivalent of O_DSYNC).