Have posixshm_test ask the kernel for the page size

[andrew]

Summary

It may be dynamic so we can't rely on PAGE_SIZE being present or
correct.

[andrew]

tests/sys/posixshm/posixshm_test.c

1127

This one doesn't look correct for non-4k pages.

[markj]

tests/sys/posixshm/posixshm_test.c

1207–1208

Why not use ps[0] here as is done elsewhere?

[markj]

tests/sys/posixshm/posixshm_test.c

1127

Yeah, looks like it's supposed to be 1 << PAGE_SHIFT.

[andrew]

• Remove a call to getpagesize when we already have it's result in the ps array
• Use page_size as the offset in the Aligned fspacectl(fd, SPACECTL_DEALLOC, ...) to OFF_MAX test

[imp]

One possible nit

tests/sys/posixshm/posixshm_test.c

1193

This weakens the test. Shouldn't it be == page_size and make page_size a global?

[andrew]

tests/sys/posixshm/posixshm_test.c

1193

It's more of a sanity check than a test.

[kib]

BTW, should PAGE_SIZE be the least guaranteed page size? In other words, should some test ensure _Static_assert(PAGE_SIZE <= ps[0]). Also perhaps something about ps[0] being multiple of PAGE_SIZE worth to be guaranteed?

[andrew]

PAGE_SIZE is passed into mmap so would need to be the maximum value. I've created D34980 to add a test that the values we get from getpagesize and getpagesizes are what we expect.