thr3ads.net - llvm dev - [LLVMdev] Does current LLVM target-independent code generator supports my strange chip? [Nov 2008]

If this information is useful, please help other people find it:
Share via:

Wei

2008-Nov-17 14:36 UTC

[LLVMdev] Does current LLVM target-independent code generator supports my strange chip?

I have a very strange and complicate H/W platform.
It has many registers in one format.
The register format is:

------------------------------
----------------------------------------------------------------------------------------
|         24-bit                |             24-bit            |
24-bit               |              24-bit            |
----------------------------------------------------------------------------------------------------------------------
              a
b                               c                                       d

There are 4 channels in a register, and each channel contains 24-bit, hence,
there are total 96-bit in 'one' register.
You can store a 24-bit integer or a s7.16 floating-point data into each
channel.
You can name each channel 'a', 'b', 'c', 'd'.

Here is an example of the operation in this H/W platform:

            ADD      R3.ab, R1.abab, R2.bbaa

it means

           Add 'abab' channel of R1 and 'bbaa' channel of R2,
and put the
result into the 'ab' channel of R3.

It's complicate.
Imagine a non-existed temp register named 'Rt1', the content of its
'a','b','c','d' channel are got from
'a','b','a','b' channel of R1,
and imagine another non-existed temp register named 'Rt2', the content
of
its 'a','b','c','d' channel are got from
'b','b','a','a' channel of R2.
and then add Rt1 & Rt2, put the result to R3
this means
the 'a' channel of R3 will be equal to the 'a' channel of Rt1
plus the 'a'
channel of Rt2, (i.e. 'a' from R1 + 'b' from R2, because
R1.'a'bab and
R2.'b'baa)
the 'b' channel of R3 will be equal to the 'b' channel of Rt1
plus the 'b'
channel of Rt2, (i.e. 'b' from R1 + 'b' from R2, because
R1.a'b'ab and
R2.b'b'aa)
the 'c' channel of R3 will be untouched, the value of the 'c'
channel of Rt1
plus the 'c' channel of Rt2 (i.e. 'a' from R1 + 'a' from
R2, because
R1.ab'a'b and R2.bb'a'a) will be lost.
the 'd' channel of R3 will be untouched, too. The value of the
'd' channel
of Rt1 plus the 'd' channel of Rt2 (i.e. 'b' from R1 +
'a' from R2, because
R1.aba'b' and R2.bba'a') will be lost, too.

I don't know whether I can set the 'type' of such register using a
llvm::MVT::SimpleValueType?
According the LLVM doc & LLVM source codes, I think llvm::MVT::v8i8, v2f32,
etc is used to represent register for SIMD instructions.
I don't think the operations in my platform are SIMD instructions.
However, I can not find any llvm::MVT::SimpleValueType which can represents
a 96-bit register.

Thus, my question is:

1) Does current LLVM backend supports this H/W?
2) If yes, how can I write the type of the register class in my .td file?

(Which value should I fill in the following 'XXX' ?)
def TempRegs : RegisterClass<"MFLEXG", [XXX], 32, [R0, R1, R2, R3,
R4, R5,
R6, R7, R8, R9,
                                                   R10, R11, R12, R13, R14,
R15, R16, R17, R18, R19,
                                                   R20, R21, R22, R23, R24,
R25, R26, R27, R28, R29,
                                                   R30, R31]> {
}

3) If not, does this means I have to write the whole LLVM backend based on
the basic llvm::TargetMachine & llvm::TargetData, just like what CBackend
does?


--------------------------------------------------------
Wei Hu
http://www.csie.ntu.edu.tw/~r88052/
<http://www.csie.ntu.edu.tw/%7Er88052/>
http://wei-hu-tw.blogspot.com/
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20081117/704aa84f/attachment.html>

Evan Cheng

2008-Nov-18 18:01 UTC

head link

[LLVMdev] Does current LLVM target-independent code generator supports my strange chip?

Why not model each channel as a separate physical register?

Evan

On Nov 17, 2008, at 6:36 AM, Wei wrote:
> I have a very strange and complicate H/W platform.
> It has many registers in one format.
> The register format is:
>
> ------------------------------
>
----------------------------------------------------------------------------------------
> |         24-bit                |             24-bit             
> |           24-bit               |              24-bit            |
>
----------------------------------------------------------------------------------------------------------------------
>               a                                        
> b                                
> c                                       d
>
> There are 4 channels in a register, and each channel contains 24- 
> bit, hence, there are total 96-bit in 'one' register.
> You can store a 24-bit integer or a s7.16 floating-point data into  
> each channel.
> You can name each channel 'a', 'b', 'c',
'd'.
>
> Here is an example of the operation in this H/W platform:
>
>             ADD      R3.ab, R1.abab, R2.bbaa
>
> it means
>
>            Add 'abab' channel of R1 and 'bbaa' channel of
R2, and
> put the result into the 'ab' channel of R3.
>
> It's complicate.
> Imagine a non-existed temp register named 'Rt1', the content of its
> 'a','b','c','d' channel are got from
'a','b','a','b' channel of R1,
> and imagine another non-existed temp register named 'Rt2', the  
> content of its 'a','b','c','d' channel are
got from 'b','b','a','a'
> channel of R2.
> and then add Rt1 & Rt2, put the result to R3
> this means
> the 'a' channel of R3 will be equal to the 'a' channel of
Rt1 plus
> the 'a' channel of Rt2, (i.e. 'a' from R1 + 'b'
from R2, because
> R1.'a'bab and R2.'b'baa)
> the 'b' channel of R3 will be equal to the 'b' channel of
Rt1 plus
> the 'b' channel of Rt2, (i.e. 'b' from R1 + 'b'
from R2, because
> R1.a'b'ab and R2.b'b'aa)
> the 'c' channel of R3 will be untouched, the value of the
'c'
> channel of Rt1 plus the 'c' channel of Rt2 (i.e. 'a' from
R1 + 'a'
> from R2, because R1.ab'a'b and R2.bb'a'a) will be lost.
> the 'd' channel of R3 will be untouched, too. The value of the
'd'
> channel of Rt1 plus the 'd' channel of Rt2 (i.e. 'b' from
R1 + 'a'
> from R2, because R1.aba'b' and R2.bba'a') will be lost,
too.
>
> I don't know whether I can set the 'type' of such register
using a
> llvm::MVT::SimpleValueType?
> According the LLVM doc & LLVM source codes, I think llvm::MVT::v8i8,  
> v2f32, etc is used to represent register for SIMD instructions.
> I don't think the operations in my platform are SIMD instructions.
> However, I can not find any llvm::MVT::SimpleValueType which can  
> represents a 96-bit register.
>
> Thus, my question is:
>
> 1) Does current LLVM backend supports this H/W?
> 2) If yes, how can I write the type of the register class in my .td  
> file?
>
> (Which value should I fill in the following 'XXX' ?)
> def TempRegs : RegisterClass<"MFLEXG", [XXX], 32, [R0, R1, R2,
R3,
> R4, R5, R6, R7, R8, R9,
>                                                    R10, R11, R12,  
> R13, R14, R15, R16, R17, R18, R19,
>                                                    R20, R21, R22,  
> R23, R24, R25, R26, R27, R28, R29,
>                                                    R30, R31]> {
> }
>
> 3) If not, does this means I have to write the whole LLVM backend  
> based on the basic llvm::TargetMachine & llvm::TargetData, just like  
> what CBackend does?
>
>
> --------------------------------------------------------
> Wei Hu
> http://www.csie.ntu.edu.tw/~r88052/
> http://wei-hu-tw.blogspot.com/
>
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20081118/84549364/attachment.html>

Wei

2008-Nov-20 12:46 UTC

head link

[LLVMdev] Does current LLVM target-independent code generator supports my strange chip?

Because each channel contains 24-bit, so.. what is the
llvm::SimpleValueType I should use for each channel?
the current llvm::SimpleValueType contains i1, i8, i16, i32, i64, f32,
f64, f80, none of them are fit one channel (24-bit).

I think I can use i32 or f32 to represent each 24-bit channel, if the
runtime result of some machine instructions exceeds 23-bit (1 bit is
for sign), then it is an overflow.
Is it correct to claim that the programmers needs to revise his
program to fix this problem?
Am I right or wrong about this thought?

If there is a chip, whose registers are 24-bit long, and you have to
compile C/C++ programs on it.
How would you represent the following statement?

int a = 3;
(Programmers think sizeof(int) = 4)

Wei.

On Nov 19, 2:01 am, Evan Cheng <evan.ch... at apple.com>
wrote:> Why not model each channel as a separate physical register?
>
> Evan
>
> On Nov 17, 2008, at 6:36 AM, Wei wrote:
>
> > I have a very strange and complicate H/W platform.
> > It has many registers in one format.
> > The register format is:
>
> > ------------------------------
> >
----------------------------------------------------------------------------------------
> > |         24-bit                |             24-bit            
> > |           24-bit               |              24-bit            |
> >
----------------------------------------------------------------------------------------------------------------------
> >               a                                        
> > b                                
> > c                                       d
>
> > There are 4 channels in a register, and each channel contains 24-
> > bit, hence, there are total 96-bit in 'one' register.
> > You can store a 24-bit integer or a s7.16 floating-point data into  
> > each channel.
> > You can name each channel 'a', 'b', 'c',
'd'.
>
> > Here is an example of the operation in this H/W platform:
>
> >             ADD      R3.ab, R1.abab, R2.bbaa
>
> > it means
>
> >            Add 'abab' channel of R1 and 'bbaa' channel
of R2, and  
> > put the result into the 'ab' channel of R3.
>
> > It's complicate.
> > Imagine a non-existed temp register named 'Rt1', the content
of its  
> > 'a','b','c','d' channel are got from
'a','b','a','b' channel of R1,
> > and imagine another non-existed temp register named 'Rt2', the
 
> > content of its 'a','b','c','d' channel
are got from 'b','b','a','a'  
> > channel of R2.
> > and then add Rt1 & Rt2, put the result to R3
> > this means
> > the 'a' channel of R3 will be equal to the 'a' channel
of Rt1 plus  
> > the 'a' channel of Rt2, (i.e. 'a' from R1 +
'b' from R2, because  
> > R1.'a'bab and R2.'b'baa)
> > the 'b' channel of R3 will be equal to the 'b' channel
of Rt1 plus  
> > the 'b' channel of Rt2, (i.e. 'b' from R1 +
'b' from R2, because  
> > R1.a'b'ab and R2.b'b'aa)
> > the 'c' channel of R3 will be untouched, the value of the
'c'  
> > channel of Rt1 plus the 'c' channel of Rt2 (i.e. 'a'
from R1 + 'a'  
> > from R2, because R1.ab'a'b and R2.bb'a'a) will be
lost.
> > the 'd' channel of R3 will be untouched, too. The value of the
'd'  
> > channel of Rt1 plus the 'd' channel of Rt2 (i.e. 'b'
from R1 + 'a'  
> > from R2, because R1.aba'b' and R2.bba'a') will be
lost, too.
>
> > I don't know whether I can set the 'type' of such register
using a  
> > llvm::MVT::SimpleValueType?
> > According the LLVM doc & LLVM source codes, I think
llvm::MVT::v8i8,  
> > v2f32, etc is used to represent register for SIMD instructions.
> > I don't think the operations in my platform are SIMD instructions.
> > However, I can not find any llvm::MVT::SimpleValueType which can  
> > represents a 96-bit register.
>
> > Thus, my question is:
>
> > 1) Does current LLVM backend supports this H/W?
> > 2) If yes, how can I write the type of the register class in my .td  
> > file?
>
> > (Which value should I fill in the following 'XXX' ?)
> > def TempRegs : RegisterClass<"MFLEXG", [XXX], 32, [R0,
R1, R2, R3,  
> > R4, R5, R6, R7, R8, R9,
> >                                                    R10, R11, R12,  
> > R13, R14, R15, R16, R17, R18, R19,
> >                                                    R20, R21, R22,  
> > R23, R24, R25, R26, R27, R28, R29,
> >                                                    R30, R31]> {
> > }
>
> > 3) If not, does this means I have to write the whole LLVM backend  
> > based on the basic llvm::TargetMachine & llvm::TargetData, just
like  
> > what CBackend does?
>
> > --------------------------------------------------------
> > Wei Hu
> >http://www.csie.ntu.edu.tw/~r88052/
> >http://wei-hu-tw.blogspot.com/
>
> > _______________________________________________
> > LLVM Developers mailing list
> > LLVM... at cs.uiuc.edu        http://llvm.cs.uiuc.edu
> >http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
>
>
>
> _______________________________________________
> LLVM Developers mailing list
> LLVM... at cs.uiuc.edu      
 http://llvm.cs.uiuc.eduhttp://lists.cs.uiuc.edu/mailman/listinfo/llvmdev

Reasonably Related Threads

Search for more reasonably related threads

llvm dev - Nov 2008 - [LLVMdev] Does current LLVM target-independent code generator supports my strange chip?

[LLVMdev] Does current LLVM target-independent code generator supports my strange chip?

[LLVMdev] Does current LLVM target-independent code generator supports my strange chip?

[LLVMdev] Does current LLVM target-independent code generator supports my strange chip?

Reasonably Related Threads