refactor: move definition of time_t in uclibc

[dybucc]

Description

This PR addresses bit-width representation issues under uClibc for the time_t type.

Following the definition in upstream uclibc-ng (the maintained fork of uClibc,) this type should be
defined in terms of either an i64 or a c_long. This depends on a macro, __UCLIBC_USE_TIME64__,
that makes the latter one available in all systems, irrespective of machine word size or
implementation memory model. This type is declared under libc/sysdeps/linux/common/bits/types.h,
in terms of other types in libc/sysdeps/linux/common/bits/typesizes.h, with no alternative
definitions for other platforms. The macro is not directly set by the library user, but rather
issued at compile time as part of the make configuration scripts that the library provides.

Yet in the libc crate, time_t under the linux/uclibc module has different definitions for all
four of MIPS32, MIPS64, arm and x86_64. This patch changes that such that the definition is shared
by all targets using uClibc, at the top-level uclibc/mod.rs module instead. This does not, at
present, affect L4RE.

Tests are still pending on most platforms.

Sources

• Upstream uclibc-ng sources showing the types.h header with the definition of the type that
  time_t is defined as and the fragment from which that "nested" defintion stems from. Note the
  comment above the latter source below, where some developer left a note on the explicit GCC
  "bit-width configurations" uclibc-ng supports.

  https://gogs.waldemar-brodkorb.de/oss/uclibc-ng/src/master/libc/sysdeps/linux/common/bits/types.h#L152
  https://gogs.waldemar-brodkorb.de/oss/uclibc-ng/src/master/libc/sysdeps/linux/common/bits/types.h#L102-L128

• Upstream uclibc-ng sources showing the use of the __UCLIBC_USE_TIME64__ macro to ensure the type
  of the type with which time_t is ultimately defined is either 64 bits or a C long.

  https://gogs.waldemar-brodkorb.de/oss/uclibc-ng/src/master/libc/sysdeps/linux/common/bits/typesizes.h#L50-L54

Checklist

• Relevant tests in libc-test/semver have been updated
• No placeholder or unstable values like *LAST or *MAX are
  included (see #3131)
• Tested locally (cd libc-test && cargo test --target mytarget);
  especially relevant for platforms that may not be checked in CI

[dybucc]

CI is failing for reasons unrelated to the changes introduced in this PR. A rerun should do it.