- User Since
- Apr 30 2013, 5:34 PM (350 w, 4 d)
Also, how does this patch compare to D72554 ?
The changes overall looks fine to me, but this patch does not seem enough to build with ENABLE_SHARED_LIB right? We still have a cyclic dependency to break first I believe.
Can you expand on the original reason for the revert? What has been fix in the patch?
Do you need someone to land this or do you have commit access now @schweitz ?
Fri, Jan 17
Thu, Jan 16
I think River fixed it already in D72822 ?
Are there other LLVM subproject who inject themselves into the global llvm shared library? Should we have a libMLIR.so instead?
Can you rebase the patch?
Wed, Jan 15
Can you remove it from the LinalgToSPIRV dependency? It shouldn't be needed there (and this is what masked the issue I believe)
Tue, Jan 14
I'd rather use relative links if you don't mind, it make it easier to work locally with a copy of the website.
Mon, Jan 13
Thanks for the contribution.
Sun, Jan 12
Currently, MLIR itself cannot be compiled as a shared library because of the cyclic dependency among MLIR components.
Sat, Jan 11
Why does clang need constexpr to diagnose the unused here? Can't the lack of side-effects be inferred without?
Thu, Jan 9
LGTM, please wait for the others that commented here to approve as well.
Wed, Jan 8
Sun, Jan 5
Sat, Dec 28
Fri, Dec 27
What is the relationship between this and the RFC?
I don't keep the Reviewed by:, first because not everyone on the line may have approved the revision so it isn't a signal I trust, second because I have yet to see a use for this, and also because we never added this information with the email reviewer.
@bondhugula : in general we strip the Phabricator metadata from the commit messages to keep only the last line "Differential Revision: "
By the way: arc patch honors the commit author when the patch is created with arcanist. This revision was uploaded in the web interface this is why you didn't get it automatically @ftynse
It is the responsibility of the person landing the patch to properly include the author in the patch, if necessary asking the author in the reviews for how they want it to appear.
Thu, Dec 26
LGTM (but wait for the others contributors to the email threads, especially during this holidays season).
(never mind this is an old revision, it just happen to popup in my inbox because of an unrelated update to it)
Ideally the codegen opt level should be controlled by an IR-level attribute based on the compile-time opt level, but that hasn't been implemented yet.
Wed, Dec 25
Can you please join this group as well: https://reviews.llvm.org/project/members/78/ ; it'll enable CI testing on your revisions
Tue, Dec 24
Do you need someone to land this for you?