This was introduced in changeset e653f8f65a
where hard coded size types were replaced with an indirection. This corrects
a case where a 'uint64_t' was replaced with a '%s', but there is no corresponding
'TYPES[64]' in the argument list, resulting in too few arguments when processing
the string
There are now separate libs for benchmark, assert, printf, putchar
start/stop:
libs/libsel4benchmark
libs/libsel4assert
libs/libsel4printf
libs/libsel4putchar
libs/libsel4startstop
The primary changes are introducing sel4/sel4.h and removing std* types
plus porting assert and IO code from the kernel to libsel4assert,
libsel4printf, libsel4putchar.
This means the code within libsel4 and the newlibs do not overload any
typical libc entities. Instead the libraries use types like
seL4_Uint32 ... instead of uint32_t. And printf is now seL4_Printf and
assert is seL4_Assert ....
Finally, the only file modified that effects kernel code is
kernel/tools/bitfield_gen.py. It needed to be modified as it generates
files for both kernel and user space. And for user space the generated code
(types_gen.h) needed to use the new types and asserts. The changes should
not change what is generated for the kernel and I did a comparison of
kernel_final.{c|s} before and after my change and the only differences
were time stamps.
Bug: #15 Streamline kernel/libsel4 and remove its libc dependencies
IA32 is 32bit version of the x86 architecture. Whilst only IA32
is supported, much of the code is generic x86. Using a generic
x86 architecture will aid in future 64bit support
Without a name, this struct (and its typedef) comes out as an anonymous type
when imported into Isabelle. This makes it difficult to automate proofs of
functions that use this type.
See also JIRA VER-434
This commit involves the PAE paging itself, refactoring the
vspace code to extra what is common between PAE and 32bit paging,
as well as some renaming to not call the root paging structure
a PD
A C++ compiler will attempt to interpret 'return info' as an invocation of a copy
constructor. Even though the copy constructor is auto-generated (and eventually
completely eliminated due to inlining) and just does obvious member copying it
still results in an intermediate invocation that must take a reference to 'info'.
Because 'info' is a register variable it is not permissable to take a reference
to it. Manually reconstructing a new info from the .words keeps everything as
Plain Old Data
This has no effect on the code, but makes this value visible as a named
constant in Isabelle when importing C sources. It avoids us having to reference
120 as a magic number in numerous proofs.
Having the padding types as INT_MAX + 1 means C++ compilers will end up using
a signed 64 bit integer, as that is the only type that can have a value that
large, as well as the negative values our enums typically contain
Domains are unnecessarily treated explicitly as 8-bit values within boot info.
Though there are existing proof constraints that limit the maximum domain value
to 8 bits, most of the code would permit domain values up to 32 bits. The
maximum value is unnecessarily constrained in boot info, a restriction which
this commit removes.
Closes VER-341
