• Use cp -f in "commit" case

• Revise "lower_file" in "commit" case

• Revise lower_file in "basic" and "revert" case

• Update according lwhsu advices

• Update according lwhsu advices

• Update according lwhsu advices

Greetings, I am Yan-Hao Wang, a FreeBSD developer who has been a summer intern at FF. I am pleased to present my Q3 status report.

In D41468#956636, @asomers wrote:

In its current state, the "base" test doesn't pass. It seems like the expected_value variable isn't getting correctly expanded. Its value is just the name of the variable. Could you please fix that?

• Fix expected_value variable

• Remove $FreeeBSD$

• Fix BSD.tests.dist

• Use conf.sh to share common function

• Remove unnecessary code and simplify again.

• Remove unnecessary code and simplify

Use fsck -p -f to check gunion and readable device

In D41645#950316, @mckusick wrote:

Your checksum test is a good way to verify that everything worked as expected though it is possible that some sort of filesystem corruption snuck in (obviously to both copies). So you might want to add a second check by running 'fsck -p -f' on the both the gunion filesystem and the original filesystem after the gunion has been dissolved. The exit status from fsck should be zero.

Format writable_device statement remove. In general, upper device is not necessary to format.

In D41645#950245, @mckusick wrote:

I am happy to see a set of gunion tests being added. The set of tests looks complete.

In general, only the lower (readable) disk partition is formatted with a filesystem. The upper (writable) layer is used and managed by gunion. It does not cause any errors for you to format it as a filesystem, but it is not necessary to do so. Attempting to access the top layer as a filesystem after it has been used by gunion is not always going to work as gunion makes no effort to keep it consistent. In particular if you use the -o offset option, any filesystem built on the top layer will cease to be usable after gunion uses it as a top layer.

Have you verified that these tests work?

The tests are described below.

1. load:
   • Use kldstat to confirm the gunion load and unload is actually work.
2. create:
   • Look /dev/ folder to confirm the gunion create and destroy is actually work.
3. commit:
   • Use checksum to check the file is actually commit.
4. offset:
   • Create two partitions and see if we can mount the union device which use the two partitions. Test the gunion commit -o offset.
5. size:
   • Use diskinfo test gunion commit -s size.
6. secsize:
   • Use diskinfo test gunion commit -S secsize.
7. gunionname:
   • Look /dev/ folder to test gunion commit -Z gunionname.
8. revert:
   • Use checksum to check the file is actually revert.

• add one test case to test gunion commit -o offset

• format the code.

• Fix file format.

• Fix the format.
• Use number style.

• Remove unnecessary xo_finish function.
• Use number style.

• Using xo_err to replace printf.
• Use number style.

• Change usage function.
• Use EXIT_ string style.

• Use number style here.

• Using xo_error instead of 'fprintf'.
• Decide use number style here.

Using xo_error instead of fprintf.

• Using sysexits.h string in all the files.

• Use EX_ string from sysexits.h if the file include it.

In some files sysexit.h is not include (below lists), I use the number style there to avoid much change.

a. main.c
b. netisr.c

The above has sysexit.h use EX_ string.

a. common.c
b. if.c
c. nhgrp.c
d. nhops.c
e. route.c
f. route_netlink.c
f. route_netlink.c

• Use xo_error instead of fprintf.
• Decide using 1 and 0 instead of EXIT_FAILURE and EXIT_SUCESS.

Old revision.

• Fix the file foramt.

• Fix the file format.

Only simple two test cases.

1. using echo to check the environment value is correct and the environment variable is actually exist.
2. try add env and delete the env, then use printenv to check that env.

• use different way to test.

• Return non-zero status on write failuer.

Originally check the xo_finish return value can achive the POSIX standard. Howeve I use (trap "" PIPE; vmstat 2>stderr; echo $? >status | true to check this, it didn't work. I find check xo_finish return value is not enough here. Becuase xo_finish will always return 0, there is already a xo_flush call before xo_finish. So we need check xo_flush return value here.

• Remove unnecessary space line.

• Remove unnecessary space line.

• Remove unnecessary space line.

• Remove unnecessary space line.

• Remove unnecessary space line.

• Remove unnecessary space line.

• Remove unnecessary space line.

Remove unnecessary space.

Remove unnecessary space.

Remove unnecessary space.

Ignore this. Create this revision accidentally.

Update with exit(EXIT_SUCCESS) to replace return (0)

I use the output of getcwd, realpath C functions, and $PWD environment variable to ensure the output of the pwd is correct.

1. realpath -> absolute path
2. getcwd -> absolute path
3. $PWD -> logical path

update get_path.c