U
    ?h                      @   s   d dl mZ d dlmZmZ d dlmZmZ d dlm	Z	m
Z
 d dlmZmZmZmZmZmZmZmZmZ d dlmZ d dlZd dlmZmZmZ d dlZeG d	d
 d
ZedG dd deZe dkre!  dS )    )List)	dataclassfield)cudafloat32)compile_ptx_for_current_devicecompile_ptx)	cossintanexploglog10log2powtanh)truedivN)CUDATestCaseskip_on_cudasimskip_unless_cc_75c                   @   sx   e Zd ZU eedZee ed< eedZ	ee ed< eedZ
ee ed< eedZee ed< eeedddZd	S )
FastMathCriterion)default_factoryfast_expectedfast_unexpectedprec_expectedprec_unexpected)testfastprecc                    s|   | t fdd| jD  | t fdd| jD  | tfdd| jD  | tfdd| jD  d S )Nc                 3   s   | ]}| kV  qd S N .0ir   r    W/var/www/html/venv/lib/python3.8/site-packages/numba/cuda/tests/cudapy/test_fastmath.py	<genexpr>   s     z*FastMathCriterion.check.<locals>.<genexpr>c                 3   s   | ]}| kV  qd S r   r    r!   r$   r    r%   r&      s     c                 3   s   | ]}| kV  qd S r   r    r!   r   r    r%   r&      s     c                 3   s   | ]}| kV  qd S r   r    r!   r'   r    r%   r&      s     )Z
assertTrueallr   r   r   r   )selfr   r   r   r    )r   r   r%   check   s    zFastMathCriterion.checkN)__name__
__module____qualname__r   listr   r   str__annotations__r   r   r   r   r*   r    r    r    r%   r      s
   
r   z4Fastmath and PTX inspection not available on cudasimc                   @   s   e Zd Zdd ZedddZedddZdd	 Zd
d Zdd Z	e
dd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zdd Zejd d! Zd"S )#TestFastMathOptionc           
      C   sx   t j||dd|}t j||d|}|| |||| t|||dd\}}t|||d\}	}|| ||	 d S )NT)devicefastmathr2   )r   jitr*   inspect_asmr   )
r)   Zpyfuncsigr2   	criterionfastverprecverfastptx_precptxr    r    r%   _test_fast_math_common   s&         
  
z)TestFastMathOption._test_fast_math_common)r8   c                    sP    fdd} fdd}| j |td d d tfd|d | j |tfd|d d S )	Nc                    s    || d< d S Nr   r    rxopr    r%   kernel2   s    z8TestFastMathOption._test_fast_math_unary.<locals>.kernelc                    s    | S r   r    )rB   rC   r    r%   device_function5   s    zATestFastMathOption._test_fast_math_unary.<locals>.device_function   Fr2   r8   Tr>   r   )r)   rD   r8   rE   rF   r    rC   r%   _test_fast_math_unary1   s          z(TestFastMathOption._test_fast_math_unaryc                    sT    fdd} fdd}| j |td d d ttfd|d | j |ttfd|d d S )	Nc                    s    ||| d< d S r?   r    rA   rB   yrC   r    r%   rE   @   s    z9TestFastMathOption._test_fast_math_binary.<locals>.kernelc                    s
    | |S r   r    )rB   rL   rC   r    r%   r2   C   s    z9TestFastMathOption._test_fast_math_binary.<locals>.devicerG   FrH   TrI   )r)   rD   r8   rE   r2   r    rC   r%   _test_fast_math_binary?   s         z)TestFastMathOption._test_fast_math_binaryc                 C   s   |  ttdgdgd d S )Ncos.approx.ftz.f32 r   r   )rJ   r	   r   r)   r    r    r%   	test_cosfN   s    zTestFastMathOption.test_cosfc                 C   s   |  ttdgdgd d S )Nsin.approx.ftz.f32 rO   )rJ   r
   r   rP   r    r    r%   	test_sinfW   s    zTestFastMathOption.test_sinfc                 C   s    |  ttdddgdgd d S )NrR   rN   div.approx.ftz.f32 rO   )rJ   r   r   rP   r    r    r%   	test_tanf`   s    zTestFastMathOption.test_tanfc                 C   s   |  ttdgdgd d S )Ntanh.approx.f32 rO   )rJ   r   r   rP   r    r    r%   
test_tanhfj   s    zTestFastMathOption.test_tanhfc                    sL   dd  fdd}|dt dgdgdd |d	t d
dgdgdd d S )Nc                 S   s   t || d< d S r?   )r   r@   r    r    r%   tanh_kernelv   s    z>TestFastMathOption.test_tanhf_compile_ptx.<locals>.tanh_kernelc                    sT   t td d d tfd| d\}}t td d d tf| d\}}| || d S )NrG   T)r3   cc)rY   )r   r   r*   )rY   r8   r;   r<   r=   r)   rX   r    r%   tanh_common_testy   s     

zCTestFastMathOption.test_tanhf_compile_ptx.<locals>.tanh_common_test)      rV   rO   )rY   r8   )r\   r   zex2.approx.ftz.f32 zrcp.approx.ftz.f32 )r   )r)   r[   r    rZ   r%   test_tanhf_compile_ptxu   s    
z)TestFastMathOption.test_tanhf_compile_ptxc                 C   s   |  ttdgdgd d S )Nzfma.rn.f32 )r   r   )rJ   r   r   rP   r    r    r%   	test_expf   s    zTestFastMathOption.test_expfc                 C   s   |  ttddgdgd d S )Nlg2.approx.ftz.f32 Z
0f3F317218rO   )rJ   r   r   rP   r    r    r%   	test_logf   s     zTestFastMathOption.test_logfc                 C   s   |  ttddgdgd d S )Nr`   Z
0f3E9A209BrO   )rJ   r   r   rP   r    r    r%   test_log10f   s     zTestFastMathOption.test_log10fc                 C   s   |  ttdgdgd d S Nr`   rO   )rJ   r   r   rP   r    r    r%   
test_log2f   s     zTestFastMathOption.test_log2fc                 C   s   |  ttdgdgd d S rc   )rM   r   r   rP   r    r    r%   	test_powf   s     zTestFastMathOption.test_powfc              	   C   s$   |  ttdgdgdgdgd d S )NrT   
div.rn.f32)r   r   r   r   )rM   r   r   rP   r    r    r%   	test_divf   s     zTestFastMathOption.test_divfc              	   C   s   dd }t d d d t t f}tj|ddd|}tj|dd|}d}tj|tj d}| t |d|f |d	d
 W 5 Q R X z|d|f |d	d
 W n tk
r   | d Y nX d S )Nc                 S   s   || | d< d S r?   r    rK   r    r    r%   f10   s    z3TestFastMathOption.test_divf_exception.<locals>.f10rG   T)r3   debug)ri   
   )Zdtypeg      $@g        z5Divide in fastmath should not throw ZeroDivisionError)r   r   r5   npemptyZassertRaisesZeroDivisionErrorZfail)r)   rh   r7   r9   r:   ZnelemZaryr    r    r%   test_divf_exception   s    z&TestFastMathOption.test_divf_exceptionc                    s   t jddddd   fdd}td d d tf}t j|dd	|}t ||}| d
|| | d|| | d
|| | d|| d S )Nzfloat32(float32, float32)Tr4   c                 S   s   | | S r   r    )abr    r    r%   foo   s    z@TestFastMathOption.test_device_fastmath_propagation.<locals>.fooc                    s&   t d}|| jk r" ||| |< d S )NrG   )r   gridsize)Zarrvalr#   rq   r    r%   bar   s    

z@TestFastMathOption.test_device_fastmath_propagation.<locals>.barrG   )r3   zdiv.approx.f32rf   zdiv.full.f32)r   r5   r   ZassertInr6   ZassertNotIn)r)   rv   r7   r9   r:   r    ru   r%    test_device_fastmath_propagation   s    
z3TestFastMathOption.test_device_fastmath_propagationN)r+   r,   r-   r>   r   rJ   rM   rQ   rS   rU   r   rW   r^   r_   ra   rb   rd   re   rg   rn   unittestZexpectedFailurerw   r    r    r    r%   r1      s$   		


			
r1   __main__)"typingr   dataclassesr   r   Znumbar   r   Znumba.cuda.compilerr   r   mathr	   r
   r   r   r   r   r   r   r   operatorr   numpyrk   Znumba.cuda.testingr   r   r   rx   r   r1   r+   mainr    r    r    r%   <module>   s   , T