[Nasm-bugs] [Bug 3392536] 2.14.01rc1 and later: label handling bug, fails to compile jchuff-sse2.asm from libjpeg

noreply-nasm at gorcunov.org noreply-nasm at gorcunov.org
Fri Dec 14 23:22:22 PST 2018


https://bugzilla.nasm.us/show_bug.cgi?id=3392536

--- Comment #8 from Michael Murashkin <ragnarokkx at inbox.ru> ---
(In reply to H. Peter Anvin from comment #7)
> MASM isn't supported on non-Microsoft platforms, for obvious reasons.
> Additionally, I think NASM is more powerful in so many ways.
> 
> This is why we have release candidates, to give people a chance to test.
> 
> We are an open source project with very limited resources. I and Chang have
> been allowed to put in *some* of our work time because of the increasing
> importance of a portable and flexible assembler, but we still have massive
> gaps on the verification front.
> 
> Cyrill is working on the problem, but it is simply a lot of work. This one
> was a simple one-character(!) bug...
> 
> The assistance you just gave is invaluable for avoiding these kinds of
> problems. I am hoping that Cyrill's infrastructure project will let us
> improve on this front in the future, and be able to add tests other than
> micro-tests, like building jpeg-turbo and x264.

I fully and totally understand it's not the right place for this conversation
(I suppose the NASM forum would be much better), however I feel I should post a
reply to your generous comment.

> Additionally, I think NASM is more powerful in so many ways.
It really is, and I know it.

> We are an open source project with very limited resources. I and Chang have
> been allowed to put in *some* of our work time because of the increasing
> importance of a portable and flexible assembler, but we still have massive
> gaps on the verification front.
Don't get me wrong, I totally appreciate your work and your commitment.

> Cyrill is working on the problem, but it is simply a lot of work. This one
> was a simple one-character(!) bug...
Grumbler mode on.
As I mentioned above, I've been using NASM for a decade, more or less. Peeping
into its code reveals a lot of those, how should I put it, oddities all the
time. NASM happens to be unable to be built at all, to be crashing just at a
start, to be unable to produce a binary, a lot of things. And there are some of
these oddities that live there for years, nobody seems to care to fix them. I
suppose those are not bugs, if they live there for so long, those are just,
well, things that don't mess with anything due to their relevance to the Win32
part of the code. They don't brake anything for the only important group of
people using NASM, because all of these people live on Linux.
Grumbler mode off.

Again, don't get me wrong, I don't complain, I fully and totally understand the
advantages of the open source software, and I really, really like to have the
ability to get access to its code and build it myself, adjusting it to my
needs. I don't have this opportunity in the proprietary software. So, I don't
really complain, I just merely grumble, just express my dull and unnecessary
here thoughts.

-- 
You are receiving this mail because:
You are watching all bug changes.
You are on the CC list for the bug.


More information about the Nasm-bugs mailing list