In the gc toolchain, the compilers and linkers now use the
same command-line flag parsing rules as the Go flag package, a departure
from the traditional Unix flag parsing. This may affect scripts that invoke
the tool directly.
For example,
go tool 6c -Fw -Dfoo must now be written
go tool 6c -F -w -D foo.
The language allows the implementation to choose whether the int type and uint types are 32 or 64 bits. Previous Go implementations made int and uint 32 bits on all systems. Both the gc and gccgo implementations (TODO: check that gccgo does) now make int and uint 64 bits on 64-bit platforms such as AMD64/x86-64.
Among other things, this enables the allocation of slices with
more than 2 billion elements on 64-bit platforms.
Updating:
Most programs will be unaffected by this change.
Because Go does not allow implicit conversions between distinct
numeric types,
no programs will stop compiling due to this change.
However, programs that contain implicit assumptions
that int is only 32 bits may change behavior.
For example, this code prints a positive number on 64-bit systems and
a negative one on 32-bit systems:
x := ^uint32(0) // x is 0xffffffff i := int(x) // i is -1 on 32-bit systems, 0xffffffff on 64-bit fmt.Println(i)
Portable code intending 32-bit sign extension (yielding -1 on all systems) would instead say:
i := int(int32(x))
Due to the int and TODO: OTHER changes, the placement of function arguments on the stack has changed. Functions written in assembly will need to be revised at least to adjust frame pointer offsets.
The implementation now includes a built-in data race detector.
In the gc toolchain, the symbol table format has been extended to allow little-endian encoding of symbol values, and the extension is used in binaries generated by the Go 1.1 version of the gc linker. To the Go 1.0 toolchain and libraries, these new symbol tables appear empty.
Previous versions of the debug/elf package intentionally skipped over the first symbol in the ELF symbol table, since it is always an empty symbol. This symbol is no longer skipped since indexes into the symbol table returned by debug/elf, will be different to indexes into the original ELF symbol table. Any code that calls the debug/elf functions Symbols or ImportedSymbols may need to be adjusted to account for the additional symbol and the change in symbol offsets.
The protocol-specific resolvers were formerly
lax about the network name passed in. For example, although the documentation was clear
that the only valid networks for ResolveTCPAddr are "tcp",
"tcp4", and "tcp6", the Go 1.0 implementation silently accepted
any string. The Go 1.1 implementation returns an error if the network is not one of those strings.
The same is true of the other protocol-specific resolvers ResolveIPAddr, ResolveUDPAddr, and
ResolveUnixAddr.
The previous ListenUnixgram returned UDPConn as
arepresentation of the connection endpoint. The Go 1.1 implementation
returns UnixConn to allow reading and writing
with ReadFrom and WriteTo methods on
the UnixConn.
On FreeBSD, Linux, NetBSD, OS X and OpenBSD, previous versions of the time package returned times with microsecond precision. The Go 1.1 implementation of time on these systems now returns times with nanosecond precision. Code may exist that expects to be able to store such a time in an external format with only microsecond precision, read it back, and recover exactly the same time instant. In Go 1.1 the same time will not be recovered, since the external storage will have discarded nanoseconds. To address this case, there are two new methods of time.Time, Round and Truncate, that can be used to remove precision from a time before passing it to external storage.
TODO