Please take a look:
http://llvm.org/docs/ReleaseNotes.html
-Chris
Please take a look:
http://llvm.org/docs/ReleaseNotes.html
GCC4.0-based llvm-gcc front-end section:
Replace the second "in addition" with something else.
"If you can use it, llvm-gcc4 is offers significant new functionality":
remove is
"In 1.8, it will be removed, replaced with the new SPARC backend."
and replaced (no comma)
"now have initial support Darwin DWARF debugging information"
..for Darwin..
Looks good.
-Tanya
Fixed, thanks! I think the release notes are good to go, if you want to suck them into the release branch.
-Chris
Chris,
Here's my review notes:
1. In the Known problems section you identify PR656 as a known problem.
However, this bug is resolved and in January you reported that you
verified it on Solaris. So, one of two things needs to happen. Either
remove this item from the release notes, or re-open the bug and tell me
what still breaks.
2. It seems that several items from the original DRAFT email of the
release notes are not in this document. Perhaps its just a matter of
time or these were intentional omissions, or the wording changed so I
didn't recognize it. But I thought I should point them out:
* Andrew added support for a new LLVM "readcyclecounter" intrinsic, for
accessing low-level target timing interfaces.
* LLVM now supports llvm.stacksave/llvm.stackrestore intrinsics, for
proper C99 Variable Length Array support.
* Nate reimplemented post-dominator analysis using the Lengauer and
Tarjan algorithm, replacing the old iterative implementation. On one
extreme example his implementation is 40x faster than the old one
(PR681) and uses far less memory.
* Daniel Berlin contributed an ET-Forest implementation, which
replaces the old LLVM DominatorSet with a far more efficient data
structure (in both space and time).
* Andrew wrote a new "reg2mem" which transforms an LLVM function so that
there are no SSA values live across basic blocks.
* The -reassociate pass knows how to factor expressions in several ways,
e.g. turning (A*A+A*B) into (A*(A+B)) and (X+X+Y+Y) into ((X+Y) << 1)
* Saem Ghani contributed support to allow different implementations of
the abstract callgraph interface, e.g., based on pointer analysis.
* Evan added support for -fpic and -static codegen on Darwin.
* Evan added initial support for subtargets in the X86 backend,
including
a broad range of -mcpu=* values.
Reid.
Here's my review notes:
1. In the Known problems section you identify PR656 as a known problem.
However, this bug is resolved and in January you reported that you
verified it on Solaris. So, one of two things needs to happen. Either
remove this item from the release notes, or re-open the bug and tell me
what still breaks.
Oops, you're right, I removed it.
2. It seems that several items from the original DRAFT email of the
release notes are not in this document. Perhaps its just a matter of
time or these were intentional omissions, or the wording changed so I
didn't recognize it. But I thought I should point them out:
Thanks, I intentionally tried to keep the release notes "higher level" than the announcement (which is already fairly high-level). Thus a lot of things get lumped into "stuff is faster".
-Chris