Displaying 6 results from an estimated 6 matches for "fixedstack3".
Did you mean:
fixedstack
2011 Aug 06
0
[LLVMdev] How to differ from read and write operations for general stack objects
...ead>, %EFLAGS<imp-def,dead>,
%ESP<imp-use>*
* MOV32mi <fi#1>, 1, %reg0, 0, %reg0, 0*
* %ECX<def> = MOV32rm <fi#1>, 1, %reg0, 0, %reg0*
* MOV32mr <fi#0>, 1, %reg0, 0, %reg0, %ECX<kill>*
* MOV32mr <fi#3>, 1, %reg0, 0, %reg0, %EAX<kill>; mem:ST4[FixedStack3]*
It seems the ways to identify read/write are different for virtual
registers, and stack objects.
For virtual registers, we could use the MachineOperand::isDef() and
MachineOperand::isUse.
For stack objects used for argument-passing (see mem:ST4[Stack] in the
example) or return-value (see mem:ST4...
2013 Feb 04
2
[LLVMdev] Asserts in bundleWithPred() and bundleWithSucc()
...not currently in the def list for the
bundle header).
v BUNDLE %P3<imp-def>, %R29<imp-use>, %D8<imp-use,kill>,
%D9<imp-use,kill>, %R6<imp-use>
*^v STrid_indexed %R29, 80, %D8<kill>; mem:ST8[FixedStack2]
*^v STrid_indexed %R29, 72, %D9<kill>; mem:ST8[FixedStack3]
*^v %P3<def> = CMPEQri %R6, 0
*^ %R17<def> = TFR_cdnNotPt %P3<internal>, %R1
v BUNDLE %R29<imp-use>, %D10<imp-use,kill>, %R7<imp-use>, %D6<imp-use>
(next bundle).
finalizeBundle() is called with:
FirstMI == STrid_indexed %R29, 80, %D8<kill>;...
2013 Feb 04
0
[LLVMdev] Asserts in bundleWithPred() and bundleWithSucc()
...the
> bundle header).
>
>
> v BUNDLE %P3<imp-def>, %R29<imp-use>, %D8<imp-use,kill>,
> %D9<imp-use,kill>, %R6<imp-use>
> *^v STrid_indexed %R29, 80, %D8<kill>; mem:ST8[FixedStack2]
> *^v STrid_indexed %R29, 72, %D9<kill>; mem:ST8[FixedStack3]
> *^v %P3<def> = CMPEQri %R6, 0
> *^ %R17<def> = TFR_cdnNotPt %P3<internal>, %R1
> v BUNDLE %R29<imp-use>, %D10<imp-use,kill>, %R7<imp-use>, %D6<imp-use>
> (next bundle).
>
> finalizeBundle() is called with:
>
> FirstMI == ST...
2013 Feb 02
0
[LLVMdev] Asserts in bundleWithPred() and bundleWithSucc()
On Feb 1, 2013, at 3:43 PM, "Sergei Larin" <slarin at codeaurora.org> wrote:
> I have a question about the following (four) asserts recently added in
> bundleWithPred() and bundleWithSucc() (see below). What is the real danger
> of reasserting a connection even if it already exist?
The intention was to identify code that may have been converted from the old style a
2013 Feb 04
2
[LLVMdev] Asserts in bundleWithPred() and bundleWithSucc()
...t; >
> >
> > v BUNDLE %P3<imp-def>, %R29<imp-use>, %D8<imp-use,kill>,
> > %D9<imp-use,kill>, %R6<imp-use> *^v STrid_indexed %R29, 80,
> > %D8<kill>; mem:ST8[FixedStack2] *^v STrid_indexed %R29, 72,
> > %D9<kill>; mem:ST8[FixedStack3] *^v %P3<def> = CMPEQri %R6, 0
> > *^ %R17<def> = TFR_cdnNotPt %P3<internal>, %R1
> > v BUNDLE %R29<imp-use>, %D10<imp-use,kill>, %R7<imp-use>,
> > %D6<imp-use> (next bundle).
> >
> > finalizeBundle() is called with:
>...
2013 Feb 01
4
[LLVMdev] Asserts in bundleWithPred() and bundleWithSucc()
Jakob,
I have a question about the following (four) asserts recently added in
bundleWithPred() and bundleWithSucc() (see below). What is the real danger
of reasserting a connection even if it already exist? My problem with them
happens when I try to call finalizeBundle() on an existing bundle to which I
have added a new instruction. The goal - a new bundle header with liveness
abbreviation, but