Challenge For The "Expert" Tyrone

Discussion:

Add Reply

Farley Flud

2025-01-28 16:14:38 UTC

Poor tired, exhausted Tyrone. He must have spent days of futile
searching in an attempt to find a copy somewhere of my absolutely
perfect AVX-512 assembly code.

(Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!)

Of course, all of his efforts were in total vain, because no such
copy exists anywhere, except right here on C.O.L.A.

Poor tired, exhausted Tyrone (not to mention poor, dumb bastard).

(Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!)

Well, I have a challenge for the "expert" Tyrone.

I have ever so slightly modified my absolutely perfect AVX-512 code
so that it no longer will execute. Instead it will crash horribly.

The ever-so-slightly modified code follows.

Let's allow the "expert" Tyrone to discover and clearly report
the fault.

Anyone want to takes bets?

Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!

I recommend that Tyrone invest his extensive and exhaustive search
time in a search for his own stupidity.

Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!

============================================
Begin AVX-512 NASM Assembly (Modified)
============================================

BITS 64

segment .text
global _start

_start:
mov r8, data_in
mov r9, data_out
mov rbx, qword [stride]
xor rdx, rdx
mov rax, qword [N]
div rbx ; rax = quotient, rdx = remainder
load:
vmovdqa32 zmm1, zword [r8]
vmovdqa32 zword [r9], zmm1
add r8, 64 ; increment data pointers
add r9, 64
dec rax
jnz load
xor r11, r11 ; load mask, i.e. only rdx left over to load
mov r10, -1
mov rcx, rdx
shld r11, r10, cl
kmovq k1, r11;
vmovdqa32 zmm1{k1}{z}, zword [r8]
vmovdqa32 zword [r9], zmm1
exit:
xor edi,edi
mov eax,60
syscall

segment .data
align 64
N: dq 37 ;set length of block and stride
stride: dq 16
data_in: dd 16 dup (0xefbeadde) ;dummy data
dd 16 dup (0xfecaafde)
dd 5 dup (0xefbeadde)

segment .bss
alignb 64
data_out: resd 37

========================================
End AVX-512 NASM Assembly (Modified)
========================================

--
Systemd: solving all the problems that you never knew you had.

Physfitfreak

2025-01-28 16:43:16 UTC

Permalink

Post by Farley Flud
Poor tired, exhausted Tyrone. He must have spent days of futile
searching in an attempt to find a copy somewhere of my absolutely
perfect AVX-512 assembly code.
(Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!)
Of course, all of his efforts were in total vain, because no such
copy exists anywhere, except right here on C.O.L.A.
Poor tired, exhausted Tyrone (not to mention poor, dumb bastard).
(Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!)
Well, I have a challenge for the "expert" Tyrone.
I have ever so slightly modified my absolutely perfect AVX-512 code
so that it no longer will execute. Instead it will crash horribly.
The ever-so-slightly modified code follows.
Let's allow the "expert" Tyrone to discover and clearly report
the fault.
Anyone want to takes bets?
Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!
I recommend that Tyrone invest his extensive and exhaustive search
time in a search for his own stupidity.
Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!
============================================
Begin AVX-512 NASM Assembly (Modified)
============================================
BITS 64
segment .text
global _start
mov r8, data_in
mov r9, data_out
mov rbx, qword [stride]
xor rdx, rdx
mov rax, qword [N]
div rbx ; rax = quotient, rdx = remainder
vmovdqa32 zmm1, zword [r8]
vmovdqa32 zword [r9], zmm1
add r8, 64 ; increment data pointers
add r9, 64
dec rax
jnz load
xor r11, r11 ; load mask, i.e. only rdx left over to load
mov r10, -1
mov rcx, rdx
shld r11, r10, cl
kmovq k1, r11;
vmovdqa32 zmm1{k1}{z}, zword [r8]
vmovdqa32 zword [r9], zmm1
xor edi,edi
mov eax,60
syscall
segment .data
align 64
N: dq 37 ;set length of block and stride
stride: dq 16
data_in: dd 16 dup (0xefbeadde) ;dummy data
dd 16 dup (0xfecaafde)
dd 5 dup (0xefbeadde)
segment .bss
alignb 64
data_out: resd 37
========================================
End AVX-512 NASM Assembly (Modified)
========================================

No Farley, not after the advent of DeepSeek.

If I, who doesn't know _hoot_ about assembly language can find the
answer within about 10 seconds of your post, then any "Tyrone" can also.

Output from DeepSeek:

1. Incorrect Handling of Remainder:

The code calculates the quotient and remainder of N / stride using
the div instruction. The quotient (rax) is used to determine how many
full 64-byte blocks to copy, and the remainder (rdx) is used to handle
the remaining elements.

However, the code assumes that the remainder (rdx) will always be
less than or equal to 16 (the stride). If N is not a multiple of stride,
the remainder could be larger than 16, which would lead to incorrect
behavior when creating the mask for the final partial load.

2. Mask Calculation:

The mask calculation for the final partial load is incorrect. The
shld instruction is used to create a mask, but it doesn't correctly
handle the case where the remainder is zero. If the remainder is zero,
the mask should be zero, but the current code will still attempt to load
and store data, which could lead to accessing out-of-bounds memory.

3. Alignment Issues:

The data_in and data_out arrays are aligned to 64 bytes, which is
correct for AVX-512 operations. However, the code does not ensure that
the final partial load is also aligned, which could lead to performance
issues or crashes on some architectures.

Joel

2025-01-28 17:09:36 UTC

Permalink

Post by Physfitfreak

No Farley, not after the advent of DeepSeek.
If I, who doesn't know _hoot_ about assembly language can find the
answer within about 10 seconds of your post, then any "Tyrone" can also.
The code calculates the quotient and remainder of N / stride using
the div instruction. The quotient (rax) is used to determine how many
full 64-byte blocks to copy, and the remainder (rdx) is used to handle
the remaining elements.
However, the code assumes that the remainder (rdx) will always be
less than or equal to 16 (the stride). If N is not a multiple of stride,
the remainder could be larger than 16, which would lead to incorrect
behavior when creating the mask for the final partial load.
The mask calculation for the final partial load is incorrect. The
shld instruction is used to create a mask, but it doesn't correctly
handle the case where the remainder is zero. If the remainder is zero,
the mask should be zero, but the current code will still attempt to load
and store data, which could lead to accessing out-of-bounds memory.
The data_in and data_out arrays are aligned to 64 bytes, which is
correct for AVX-512 operations. However, the code does not ensure that
the final partial load is also aligned, which could lead to performance
issues or crashes on some architectures.

Ha!
--
Joel W. Crump

Amendment XIV
Section 1.

[...] No state shall make or enforce any law which shall
abridge the privileges or immunities of citizens of the
United States; nor shall any state deprive any person of
life, liberty, or property, without due process of law;
nor deny to any person within its jurisdiction the equal
protection of the laws.

Dobbs rewrites this, it is invalid precedent. States are
liable for denying needed abortions, e.g. TX.

Farley Flud

2025-01-28 17:51:37 UTC

Permalink

Post by Physfitfreak
No Farley, not after the advent of DeepSeek.
If I, who doesn't know _hoot_ about assembly language can find the
answer within about 10 seconds of your post, then any "Tyrone" can also.

Whoa! Anyone who trusts AI is a total idiot!

This shit is wrong everywhere.

Ha, ha, ha, ha, ha, ha, ha, ha, ha!

There is a sucker born every minute, and it's not just those
retards who attend circuses.

--
Systemd: solving all the problems that you never knew you had.

Joel

2025-01-28 17:59:33 UTC

Permalink

Post by Farley Flud

Whoa! Anyone who trusts AI is a total idiot!
This shit is wrong everywhere.
Ha, ha, ha, ha, ha, ha, ha, ha, ha!
There is a sucker born every minute, and it's not just those
retards who attend circuses.

You're a sucker for your own narcissistic ways, Physfitfreak seems to
have given a rebuttal to your erroneous code, and you just say "fuck
AI! HA HA HA", it's a bunch of bullshit. You're a fraud.
--
Joel W. Crump

Amendment XIV
Section 1.

[...] No state shall make or enforce any law which shall
abridge the privileges or immunities of citizens of the
United States; nor shall any state deprive any person of
life, liberty, or property, without due process of law;
nor deny to any person within its jurisdiction the equal
protection of the laws.

Dobbs rewrites this, it is invalid precedent. States are
liable for denying needed abortions, e.g. TX.

Physfitfreak

2025-01-28 18:25:21 UTC

Permalink

Post by Farley Flud

Whoa! Anyone who trusts AI is a total idiot!
This shit is wrong everywhere.
Ha, ha, ha, ha, ha, ha, ha, ha, ha!
There is a sucker born every minute, and it's not just those
retards who attend circuses.

Hmm.. Obviously I'm not in a position to refute your claim in this
particular case. But have you considered the AI that's provided to the
world by the Chinese might not have inserts in it to dumb it down like
in ChatGPT?

DeepSeek scared the hell out of the AI businesses in this country just
last week. Even stock market dived as a result. That should be for a
reason. They're getting good! And they're out of the control of AI Capones.

Time to put your bias aside and begin taking advantage of AI. Not
necessarily ChatGPT (it's been crippled into "safe AI"), but the ones
that come from other parts of the world. The one's made by Modern Humans
:) They don't cripple them. They're not Capones.

Star (in sci.phys) yesterday used DeepSeek to reverse engineer an EXE
file...! It correctly found the Windows password from the file. He
(Star) is still testing it for other hack issues, fighting other Capones.

A lot of individuals in this country (and some in Europe) are nervous as
of last week. I have the stock market dive to prove it. And these guys
are only the Capones of various businesses. The one's who were, up to
just last week, untouchables. Above law!

Farley Flud

2025-01-28 21:50:55 UTC

Permalink

Post by Physfitfreak
They're getting good!

Certainly you are joking.

The total failure with this extremely simple bit of code indicates
something that only a hapless idiot would trust.

Post by Physfitfreak
Time to put your bias aside

There is no bias. The result you presented is total bullshit.

My code is perfect, as anyone who has the skills could easily
test.

--
Systemd: solving all the problems that you never knew you had.

Physfitfreak

2025-01-28 22:21:40 UTC

Permalink

Post by Farley Flud

Post by Physfitfreak
They're getting good!

Certainly you are joking.
The total failure with this extremely simple bit of code indicates
something that only a hapless idiot would trust.

Post by Physfitfreak
Time to put your bias aside

There is no bias. The result you presented is total bullshit.
My code is perfect, as anyone who has the skills could easily
test.

Ok then that leaves it to Tyrone to prove you wrong. I think your
chances to win are high with him :)

Farley Flud

2025-01-28 22:45:31 UTC

Permalink

Post by Physfitfreak
Ok then

No, it's not OK.

You have chosen to introduce this patently falsifiable entity
known as "AI" and now you must suffer the consequences.

AI, like Microslop Winblows, will succeed mightily.

But, like Microslop Winbows, AI is essentially bullshit.

I suppose that AI will separate the supremely intelligent
(like me) from the common dregs who will lap it up like
the dumb dogs that they are.

Post by Physfitfreak
that leaves it to Tyrone to prove you wrong

Leaving it to Tyrone is like leaving the cure for cancer to
the local clerk at 7-11.

Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!

--
Systemd: solving all the problems that you never knew you had.

Joel

2025-01-28 23:01:13 UTC

Permalink

Post by Farley Flud
You have chosen to introduce this patently falsifiable entity
known as "AI" and now you must suffer the consequences.
AI, like Microslop Winblows, will succeed mightily.
But, like Microslop Winbows, AI is essentially bullshit.
I suppose that AI will separate the supremely intelligent
(like me) from the common dregs who will lap it up like
the dumb dogs that they are.

If you were really such a genius, and not a weirdo, you might realize
that AI is good chat about pet topics, but keep running a computer
with no case or SSD.
--
Joel W. Crump

Amendment XIV
Section 1.

[...] No state shall make or enforce any law which shall
abridge the privileges or immunities of citizens of the
United States; nor shall any state deprive any person of
life, liberty, or property, without due process of law;
nor deny to any person within its jurisdiction the equal
protection of the laws.

Dobbs rewrites this, it is invalid precedent. States are
liable for denying needed abortions, e.g. TX.

Physfitfreak

2025-01-29 00:06:47 UTC

Permalink

Post by Farley Flud

Post by Physfitfreak
Ok then

No, it's not OK.
You have chosen to introduce this patently falsifiable entity
known as "AI" and now you must suffer the consequences.
AI, like Microslop Winblows, will succeed mightily.
But, like Microslop Winbows, AI is essentially bullshit.
I suppose that AI will separate the supremely intelligent
(like me) from the common dregs who will lap it up like
the dumb dogs that they are.

Post by Physfitfreak
that leaves it to Tyrone to prove you wrong

Leaving it to Tyrone is like leaving the cure for cancer to
the local clerk at 7-11.
Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!

This has nothing to do with me, nor with AI. I was just pointing out
that challenging Tyrone with an assembly question will not work because
he'll seek the assistance of the new DeepSeek to answer your challenge.
You'd be wrestling with DeepSeek, not Tyrone.

DFS

2025-01-29 03:30:11 UTC

Permalink

Post by Farley Flud
My code is perfect, as anyone who has the skills could easily
test.

Why are you so extraordinarily proud of your tiny code routines that
don't do anything useful (or in most cases don't work at all)?

Tyrone

2025-02-01 05:50:33 UTC

Permalink

On Jan 28, 2025 at 11:14:38 AM EST, "Farley Flud" <***@linux.rocks> wrote:

His usual childish drivel.

But seriously. This is your idea of a "challenge"?

God you are such a rookie. You removed an Align 64 instruction in the data
segment. Doing this corrupts the stack, which is why it will crash horribly.

Try again, junior.