similar to: RFC: Disambiguate RegClass->getSize()

Ok, I've found that marking tiny live intervals as not spillable inside VirtRegAuxInfo::CalculateWeightAndHint is not playing nicely with very constrained regclasses, in my case a regclass composed of only one register. As a workaround, instead of marking them as not spillable, I've marked them with a very high spill cost and the regalloc is able to compile the function with good code

Hi everyone, I’m currently playing with the clang driver and I encounter a strange bug. It occurs when I used the getSize function from the SymbolRef class. On iOS, the number returned is not always correct (some function have the right size). Sadly, This is my code (at the end of the main function into driver.cpp) : #### ErrorOr<OwningBinary<Binary>> BinaryOrErr =

On Jan 9, 2013, at 10:46 AM, Borja Ferrer <borja.ferav at gmail.com> wrote: > Ok, I've found that marking tiny live intervals as not spillable inside VirtRegAuxInfo::CalculateWeightAndHint is not playing nicely with very constrained regclasses, in my case a regclass composed of only one register. > As a workaround, instead of marking them as not spillable, I've marked them

On Mar 30, 2012, at 3:12 PM, Benjamin Kramer <benny.kra at googlemail.com> wrote: > > On 30.03.2012, at 23:30, Jakob Stoklund Olesen wrote: > >> Is this function required? >> >> MCInstrInfo::getName(OpCode) returns the same string, and we have two copies of the string table now. >> >> Can MCInstrPrinter refer to MCInstrInfo, and why not? > >

Looking around today, it appears that TargetRegisterClass and MCRegisterClass only includes a single alignment. This is documented as being the minimum legal alignment, but it appears to often be greater than this in practice. For instance, on x86 the alignment of %ymm0 is listed as 32, not 1. Does anyone know why this is? Additionally, where are these alignments actually defined? I

On Jan 7, 2013, at 4:58 AM, Borja Ferrer <borja.ferav at gmail.com> wrote: > Hello Jakob, > > Did you get a chance to take a look into this, and if not, can you do it when you get some spare time? It's not likely I'll have time to look at this in the near future. I'd recommend you do it yourself. /jakob > 2012/12/19 Borja Ferrer <borja.ferav at gmail.com>

On Sep 22, 2009, at 4:49 PM, Talin wrote: > > // Calculate the size of the specified LLVM type. > Constant * DebugInfoBuilder::getSize(const Type * type) { > Constant * one = ConstantInt::get(Type::Int32Ty, 1); > return ConstantExpr::getPtrToInt( > ConstantExpr::getGetElementPtr( > ConstantPointerNull::get(PointerType::getUnqual(type)), >

Hello Jakob, Did you get a chance to take a look into this, and if not, can you do it when you get some spare time? Thanks! 2012/12/19 Borja Ferrer <borja.ferav at gmail.com> > We did something like this back when the register allocator couldn't split >> live ranges. >> > > Yes, I remember the isWinToJoinCrossClass() function, removed here: > >

On Tue, Sep 22, 2009 at 9:21 AM, Devang Patel <devang.patel at gmail.com>wrote: > On Tue, Sep 22, 2009 at 9:21 AM, Devang Patel <devang.patel at gmail.com> > wrote: > > On Tue, Sep 22, 2009 at 12:14 AM, Talin <viridia at gmail.com> wrote: > >> So, one feature of the late, lamented DebugInfoBuilder that I am missing > >> quite badly, and which is not

If someone is able, can you tell me if there is a better way to do this? More specifically, do I have to rewrite all of the data members stuff and extend stuff of parent class in the child class? See below. Thanks in advance! Example 1: setConstructorS3("ClassA", function(A,x) { if(missing(A))A=15; if(missing(x))x=NA; extend(Object(), "ClassA", .size = A, .x=x

On Wed, Sep 23, 2009 at 1:51 PM, Dan Gohman <gohman at apple.com> wrote: > > On Sep 22, 2009, at 4:49 PM, Talin wrote: > >> >> // Calculate the size of the specified LLVM type. >> Constant * DebugInfoBuilder::getSize(const Type * type) { >> Constant * one = ConstantInt::get(Type::Int32Ty, 1); >> return ConstantExpr::getPtrToInt( >>

On 08/31/2015 03:59 PM, Matthias Braun wrote: > Looks to me like the alignment is specified in tablegen. From Target.td: > > class RegisterClass<string namespace, list<ValueType> regTypes, int alignment, > dag regList, RegAltNameIndex idx = NoRegAltName> > > X86RegisterInfo.td: > > def VR256 : RegisterClass<"X86", [v32i8,

On Wed, Sep 23, 2009 at 2:27 PM, Talin <viridia at gmail.com> wrote: > On Wed, Sep 23, 2009 at 1:51 PM, Dan Gohman <gohman at apple.com> wrote: >> >> On Sep 22, 2009, at 4:49 PM, Talin wrote: >>> >>> // Calculate the size of the specified LLVM type. >>> Constant * DebugInfoBuilder::getSize(const Type * type) { >>>    Constant * one =

On Tue, Sep 22, 2009 at 4:49 PM, Talin <viridia at gmail.com> wrote: > > > On Tue, Sep 22, 2009 at 9:21 AM, Devang Patel <devang.patel at gmail.com> > wrote: >> >> On Tue, Sep 22, 2009 at 9:21 AM, Devang Patel <devang.patel at gmail.com> >> wrote: >> > On Tue, Sep 22, 2009 at 12:14 AM, Talin <viridia at gmail.com> wrote: >>

Hi Rafael, We will try out your patch and check to see how it will fit. You also talked about "It might be even possible to drop the requirement for the size to be known: Replace the call to AtEndOfStream by just trying to read more and checking if it failed, but that is a bit more than I wanted to do for this." That is to remove some calls to getSize()? Is there any expectation that

I have posted a patch that switches the API to one that supports this (yet non-existent functionality) earlier: https://reviews.llvm.org/D24631 The comments from that were incorporated into the following RFC. Motivation: Certain targets feature "variable-sized" registers, i.e. a situation where the register size can be configured by a hardware switch. A common instruction set

Hi, (This only concerns MISNeedChainEdge(), and is separate from D8705) I found out that the MIScheduler (pre-ra) could not handle a simple test case (test/CodeGen/SystemZ/alias-01.ll), with 16 independent load / add / stores. The buildSchedGraph() put too many edges between memory accesses, because 1) There was no implementation of areMemAccessesTriviallyDisjoint() for SystemZ. 2) Type

Dear Prof. Adve, I recently checked out the llvm using cvs update command under the llvm directory. But when I went to see the DSNode.h, there is no getPointerSize function. I checked the doxygen tree, it's there. So that means I didn't update my cvs tree correctly. Could you let me know how to update my CVS tree? Or there is a problem at somewhere else? Thanks, xiaodong On Sun, 10 Nov

1. What is the difference between DSNodeHandle and DSNode? What do the functions getLink() and getSize() do? 2. In the previous email, you mentioned that we can use DSNode::getPointerSize() to get the number of links, But I checked the doxygen documentation, there is no such member for DSNode. 3. Previously I use the following code: for( df_iterator<DSNode*> I = df_begin(pnode),

Right now we default to returning objects instead of arrays. perl/python have alternates that return tuples/lists. Do we want to %alias those so we can get arrays back or does it really matter?