similar to: [LLVMdev] Removing TargetMachine CPU auto-detection for PowerPC and SystemZ?

Hal Finkel <hfinkel at anl.gov> wrote on 23.03.2015 19:42:06: > > >Attached are draft patches that do this for X86. Similar but smaller > > >cleanups can also be done for SystemZ and PowerPC if it’s agreed > > >this is a good idea. > > > > However, this was then never implemented for SystemZ and PowerPC. > > Should we do so as well? > >

Andrew Kaylor wrote: >In general, the current "strict FP" handling stops at instruction >selection. At the MachineIR level we don't currently have a mechanism >to prevent inappropriate optimizations based on floating point >constraints, or indeed to convey such constraints to the backend. >Implicit register use modeling may provide some restriction on some

> On Jan 9, 2018, at 1:53 PM, Kaylor, Andrew via cfe-dev <cfe-dev at lists.llvm.org> wrote: > > I think we’re going to need to create a new mechanism to communicate strict FP modes to the backend. I think we need to avoid doing anything that will require re-inventing or duplicating all of the pattern matching that goes on in instruction selection (which is the reason we’re

I think we're going to need to create a new mechanism to communicate strict FP modes to the backend. I think we need to avoid doing anything that will require re-inventing or duplicating all of the pattern matching that goes on in instruction selection (which is the reason we're currently dropping that information). I'm out of my depth on this transition, but I think maybe we could

Hans Wennborg wrote: >Dear testers, > >8.0.0-rc4 was just tagged from the release_80 branch at r355690. I've tested the current branch on SystemZ without problems. I noticed I forgot to send a ReleaseNotes updates, sorry. If there's still time, here's a list of the major user-visible changes: (See attached file: systemz-releasenotes.diff) Mit freundlichen Gruessen / Best

Hi Andrew, sorry for the delay, I only now got some time to look into this a bit more. But I still have a number of questions of how to actually implement this in the back end. Looking at this bottom-up, starting with the behavior of the actual machine instructions, we have (at least on SystemZ) the following things to consider: A) Rounding mode Most FP arithmetic instructions use the

>The standard argument against trying to introduce "scope-like" mechanisms to LLVM IR is inlining; >unless you're going to prevent functions that use stricter/laxer FP rules from being inlined >into >each other (which sounds disastrous), you're going to need to communicate strictness on an >instruction-by-instruction basis. If the backend wants to handle that by

Oh, and other thing: Are you planning to attend the upcoming LLVM developer's meeting in Bristol? I thought it might be good idea to get all parties interested in this feature together in person, if we're at the same meeting anyway. So I was thinking of submitting a proposal for a BoF session on this topic ... Mit freundlichen Gruessen / Best Regards Ulrich Weigand -- Dr. Ulrich

On 9 Jan 2018 22:55, "John McCall via llvm-dev" <llvm-dev at lists.llvm.org> wrote: On Jan 9, 2018, at 3:50 PM, Kaylor, Andrew <andrew.kaylor at intel.com> wrote: >The standard argument against trying to introduce "scope-like" mechanisms to LLVM IR is inlining; >unless you're going to prevent functions that use stricter/laxer FP rules from being inlined

On Tue, Jan 09, 2018 at 06:53:51PM +0000, Kaylor, Andrew via cfe-dev wrote: > I think we're going to need to create a new mechanism to communicate > strict FP modes to the backend. I think we need to avoid doing anything > that will require re-inventing or duplicating all of the pattern > matching that goes on in instruction selection (which is the reason >

Hello, as of Jul 9, the LNT build bot on SystemZ shows red. The problem occurs in the "setup lit" test, which fails due to: Processing dependencies for LNT==0.4.2.dev0 Searching for typing Reading https://pypi.org/simple/typing/ Downloading

Hello, as discussed here in more detail: https://reviews.llvm.org/D81583 the introduction of the C++20 [[no_unique_address]] attribute exposes an ABI issue on platforms that require special handling for structs/classes that are "equivalent" to a single floating-point member (or in some cases, a "homogeneous" set of floating-point members). This is because we can now for the

Hi Ulrich, and any other llvm PowerPC/JIT users, It looks like the Numba maintainers have run in to an issue with RuntimeDyldELF's PowerPC support (See https://github.com/numba/numba/issues/2451#issuecomment-377739948 and later comments) Due to a recent change to weak symbol handling, we now always resolve to the first copy of a function that is emitted (discarding redundant weak/odr

Richard Sandiford wrote: >I'd like to step down as code owner for the SystemZ port and nominate >Ulrich Weigand to take over. Sorry for not doing this sooner. I'd be happy to take over that role. Thanks for all your work on SystemZ, Richard! Bye, Ulrich

Hi Ulrich, I am interested in knowing if the current proposals also take into account the FP_CONTRACT pragma and the ability to implement options that imply a specific value for the FLT_EVAL_METHOD macro. Additionally, I am not aware of the IR being able to represent the potentially deferred loss of precision that the C language semantics provide; in particular, applying such semantics to the

Hello, at the recent EuroLLVM developer meeting in Bristol I held a BoF session on the topic "Towards implementing #pragma STDC FENV_ACCESS". I've also had a number of follow-on discussions both on-site in Bristol and online since. This post is intended as a summary of my current understanding set of requirements and implementation details covering the overall topic. I'm

Hello, this set of patches adds SystemZ support to base LLVM. This is split up into three patches just to simplify review; they should be committed as a single change if approved. The first patch adds SystemZ support to the configure logic and support for the systemz Triple. This ought to be straightfoward. The second patch adds changes to LLVM common code required to support the SystemZ ELF

On 05/23/2018 11:06 AM, Hubert Tong via llvm-dev wrote: > Hi Ulrich, > > I am interested in knowing if the current proposals also take into > account the FP_CONTRACT pragma We should already do this (we turn relevant operations into the @llvm.fmuladd. when FP_CONTRACT is set to on during IR generation). > and the ability to implement options that imply a specific value for >

On Wed, May 23, 2018 at 12:19 PM, Hal Finkel <hfinkel at anl.gov> wrote: > > On 05/23/2018 11:06 AM, Hubert Tong via llvm-dev wrote: > > Hi Ulrich, > > I am interested in knowing if the current proposals also take into account > the FP_CONTRACT pragma > > > We should already do this (we turn relevant operations into the > @llvm.fmuladd. when FP_CONTRACT is

All, Currently the X86 backend does CPU auto-detection and subtarget feature detection when the TargetMachine is created if no explicit CPU was specified. It's counterintuitive for low level tools like ‘llc’ to do this, as it means the same .ll file compiled on heterogenous machines generates different results from the same ‘llc’ command line. It is still useful to be able to opt-in to such