Re: [racket-dev] Why is collections is failing pkg-build?

[Alexis King]

> On Oct 14, 2016, at 1:49 PM, Matthew Flatt  wrote:
> 
> I'll work on this as soon as I can. Meanwhile, you might have intended
> for "functional-lib" to depend on "collections-lib", and it's possible
> that change avoid the bug in the dependency calculation.

Wonderful, thank you for the explanation! No rush on fixing it,
since you are indeed right: I only recently split my collections
package into lib/doc/test, and I apparently forgot to update
functional to depend on the split out variants, which triggered the
issue. I’ve fixed functional-lib now, so if that works fine, it
should be fixed for the time being, anyway.

Thanks again,
Alexis

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-dev+unsubscr...@googlegroups.com.
To post to this group, send email to racket-dev@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/racket-dev/718A9546-DF32-4158-9239-F1BC2E74E651%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-dev] Re: Why is collections is failing pkg-build?

[Matthew Flatt]

I think it's a problem with the package server's calculation of mutual
dependencies. It generates the build plan

 functional-doc
 collections-doc
 collections-lib collections functional-lib

where the third step there correctly identifies three packages as
needing to be built together, because they're mutually dependent.
However, "functional-doc" and "collections-doc" should be in the same
mutually dependent set --- because "functional-doc" depends on
"functional-lib", which depends on "collections", which depends on
collections-doc".

I'll work on this as soon as I can. Meanwhile, you might have intended
for "functional-lib" to depend on "collections-lib", and it's possible
that change avoid the bug in the dependency calculation.

At Fri, 14 Oct 2016 10:59:38 -0700, Alexis King wrote:
> Can anyone answer this? I’d really like to get my packages’
> documentation indexed again, but I really have no idea what’s wrong.
> 
> > On Oct 6, 2016, at 9:15 PM, Alexis King  wrote:
> > 
> > I recently split up my collections package into the usual lib, doc, and
> > test packages, and I noticed today that pkg-build seems to be barfing on
> > it in such a way that is causing its dependencies to have issues, too.
> > Unfortunately, I can’t seem to figure out what’s wrong. If I look at
> > http://pkg-build.racket-lang.org, it claims that the collections package
> > has “conflicts in dependency”, but clicking on the link gives me
> > information that doesn’t seem to make much sense. Furthermore, all of
> > the collections-{lib,doc,test} packages say that “install both succeeds
> > and fails”, which manages to be even more confusing.
> > 
> > What am I missing here? How can I get these to build properly? The build
> > failure is also affecting downstream dependents, such as my functional
> > and megaparsack libraries.
> > 
> > Alexis
> > 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Racket Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to racket-dev+unsubscr...@googlegroups.com.
> To post to this group, send email to racket-dev@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/racket-dev/B7D9731F-E236-4023-8996-C5055A761E
> 25%40gmail.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-dev+unsubscr...@googlegroups.com.
To post to this group, send email to racket-dev@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/racket-dev/20161014204944.3493A6501C9%40mail-svr1.cs.utah.edu.
For more options, visit https://groups.google.com/d/optout.

[racket-dev] Re: Why is collections is failing pkg-build?

[Alexis King]

Can anyone answer this? I’d really like to get my packages’
documentation indexed again, but I really have no idea what’s wrong.

> On Oct 6, 2016, at 9:15 PM, Alexis King  wrote:
> 
> I recently split up my collections package into the usual lib, doc, and
> test packages, and I noticed today that pkg-build seems to be barfing on
> it in such a way that is causing its dependencies to have issues, too.
> Unfortunately, I can’t seem to figure out what’s wrong. If I look at
> http://pkg-build.racket-lang.org, it claims that the collections package
> has “conflicts in dependency”, but clicking on the link gives me
> information that doesn’t seem to make much sense. Furthermore, all of
> the collections-{lib,doc,test} packages say that “install both succeeds
> and fails”, which manages to be even more confusing.
> 
> What am I missing here? How can I get these to build properly? The build
> failure is also affecting downstream dependents, such as my functional
> and megaparsack libraries.
> 
> Alexis
> 

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-dev+unsubscr...@googlegroups.com.
To post to this group, send email to racket-dev@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/racket-dev/B7D9731F-E236-4023-8996-C5055A761E25%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-dev] Inconsistent behavior when doing bitwise arithmetic

[Gustavo Massaccesi]

> Those numbers look like pointers interpreted as fixnums

Yes.

The optimizer thought that the result of (bitwise-and num -2) was a
fixnum, so it changed
  (bitwise-ior (bitwise-and num -2) 0)
to
  (unsafe-fxior (bitwise-and num -2) 0)


All the pointers are even (actually a multiple of 4 in 32 bits and a
multiple of 8 in 64 bits). Fixnum are implemented as "pointers" to an
odd address, the fixnum n is encoded as 2*n+1.

I.E. If the "pointer" is even then it's a pointer to a real object. If
the "pointer" is odd then it's a fixnum.

In this case, the result of (bitwise-and num -2) was a pointer to a
bignum, but unsafe-fxior interpreted it as fixnum without checking the
parity, so what you see is the address of the bignum divided by 2.

Each iteration produces a new bignum, so you will usually get results
in an increasing order, as the allocator creates new objects.

Gustavo


On Fri, Oct 14, 2016 at 9:39 AM, Vincent St-Amour
 wrote:
> Those numbers look like pointers interpreted as fixnums.
>
> So my guess as to why they differ so much is that between each
> operation, you call `printf`, which allocates enough to claim the
> addresses between, e.g., 70112357026588 and 70112357038692. So that next
> time around the loop, the result of the `bitwise-and` lives at
> 70112357038692.
>
> Just a guess, though.
>
> Vincent
>
>
>
> On Fri, 14 Oct 2016 01:11:56 -0500,
> peter.sama...@gmail.com wrote:
>>
>> Do you have any guess why the resulting numbers vary so much?
>>
>> On Thursday, October 13, 2016 at 7:20:47 PM UTC+2, mflatt wrote:
>>
>> Thanks for the report!
>>
>> This is a bug in the optimizer's handling of `bitwise-and`, where I
>> made it assume a fixnum result if either argument is a fixnum.
>> That's
>> obviously not correct if the fixnum is negative. I'll push a repair.
>>
>> The difference you see when running in a module is because the
>> optimizer constant-folds the calculation, so there's no `bitwise-and
>> `
>> call left to make incorrect assumptions about.
>>
>> At Thu, 13 Oct 2016 09:17:52 -0700 (PDT), peter@gmail.com wrote:
>> > Hi all,
>> >
>> > I get a weird behavior when using bitwise-ior and bitwise-and with
>> large
>> > numbers. Tested on 2 machines (racket 6.6, Ubuntu 16.04 and
>> 14.04):
>> >
>> > Here is the test example:
>> >
>> > #lang racket
>> > (define num #x) ;; remove one f, and the results
>> are fine
>> > in both cases
>> > (for ([i 5])
>> > (printf "~a~n" (bitwise-ior (bitwise-and num -2) 0)))
>> >
>> >
>> > When run from DrRacket using ctr+r, it works properly:
>> > 18446744073709551614
>> > 18446744073709551614
>> > 18446744073709551614
>> > 18446744073709551614
>> > 18446744073709551614
>> >
>> > When pasted into DrRacket's REPL (or run from command line via
>> "racket
>> > random-bug.rkt"), the code does not only produce wrong result, but
>> also
>> > seems to be counting something :)
>> > 70112357026588
>> > 70112357038692
>> > 70112357043588
>> > 70112357048576
>> > 70112357053344
>> >
>> > Removing one "f" from the test number results in correct behavior
>> in both
>> > cases.
>> > Very large numbers are are reduced to the same range (around 54
>> bits).
>> > Replacing the expression (bitwise-and num -2) by the corresponding
>> number
>> > it computes results in correct behavior.
>> >
>> > Cheers,
>> > Peter
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> Groups
>> > "Racket Developers" group.
>> > To unsubscribe from this group and stop receiving emails from it,
>> send an
>> > email to racket-dev+...@googlegroups.com.
>> > To post to this group, send email to racke...@googlegroups.com.
>> > To view this discussion on the web visit
>> >
>> 
>> https://groups.google.com/d/msgid/racket-dev/d0310d90-0d88-4902-b430-0069b310d7
>>
>> > a3%40googlegroups.com.
>> > For more options, visit https://groups.google.com/d/optout.
>>
>>
>> --
>> You received this message because you are subscribed to the Google
>> Groups "Racket Developers" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an email to racket-dev+unsubscr...@googlegroups.com.
>> To post to this group, send email to racket-dev@googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/racket-dev/c4b51560-736a-4a19-9b9f-0b28f4de66cb%40googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>
> --
> You received this message because you are subscribed to the Google Groups 
> "Racket Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to racket-dev+unsubscr...@googlegroups.com.
> To post to this group, send 

Re: [racket-dev] Inconsistent behavior when doing bitwise arithmetic

[Vincent St-Amour]

Those numbers look like pointers interpreted as fixnums.

So my guess as to why they differ so much is that between each
operation, you call `printf`, which allocates enough to claim the
addresses between, e.g., 70112357026588 and 70112357038692. So that next
time around the loop, the result of the `bitwise-and` lives at
70112357038692.

Just a guess, though.

Vincent



On Fri, 14 Oct 2016 01:11:56 -0500,
peter.sama...@gmail.com wrote:
> 
> Do you have any guess why the resulting numbers vary so much?
> 
> On Thursday, October 13, 2016 at 7:20:47 PM UTC+2, mflatt wrote:
> 
> Thanks for the report! 
> 
> This is a bug in the optimizer's handling of `bitwise-and`, where I 
> made it assume a fixnum result if either argument is a fixnum.
> That's 
> obviously not correct if the fixnum is negative. I'll push a repair. 
> 
> The difference you see when running in a module is because the 
> optimizer constant-folds the calculation, so there's no `bitwise-and
> ` 
> call left to make incorrect assumptions about. 
> 
> At Thu, 13 Oct 2016 09:17:52 -0700 (PDT), peter@gmail.com wrote: 
> > Hi all, 
> > 
> > I get a weird behavior when using bitwise-ior and bitwise-and with
> large 
> > numbers. Tested on 2 machines (racket 6.6, Ubuntu 16.04 and
> 14.04): 
> > 
> > Here is the test example: 
> > 
> > #lang racket 
> > (define num #x) ;; remove one f, and the results
> are fine 
> > in both cases 
> > (for ([i 5]) 
> > (printf "~a~n" (bitwise-ior (bitwise-and num -2) 0))) 
> > 
> > 
> > When run from DrRacket using ctr+r, it works properly: 
> > 18446744073709551614 
> > 18446744073709551614 
> > 18446744073709551614 
> > 18446744073709551614 
> > 18446744073709551614 
> > 
> > When pasted into DrRacket's REPL (or run from command line via
> "racket 
> > random-bug.rkt"), the code does not only produce wrong result, but
> also 
> > seems to be counting something :) 
> > 70112357026588 
> > 70112357038692 
> > 70112357043588 
> > 70112357048576 
> > 70112357053344 
> > 
> > Removing one "f" from the test number results in correct behavior
> in both 
> > cases. 
> > Very large numbers are are reduced to the same range (around 54
> bits). 
> > Replacing the expression (bitwise-and num -2) by the corresponding
> number 
> > it computes results in correct behavior. 
> > 
> > Cheers, 
> > Peter 
> > 
> > -- 
> > You received this message because you are subscribed to the Google
> Groups 
> > "Racket Developers" group. 
> > To unsubscribe from this group and stop receiving emails from it,
> send an 
> > email to racket-dev+...@googlegroups.com. 
> > To post to this group, send email to racke...@googlegroups.com. 
> > To view this discussion on the web visit 
> >
> 
> https://groups.google.com/d/msgid/racket-dev/d0310d90-0d88-4902-b430-0069b310d7
>
> > a3%40googlegroups.com. 
> > For more options, visit https://groups.google.com/d/optout. 
> 
> 
> -- 
> You received this message because you are subscribed to the Google
> Groups "Racket Developers" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to racket-dev+unsubscr...@googlegroups.com.
> To post to this group, send email to racket-dev@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/racket-dev/c4b51560-736a-4a19-9b9f-0b28f4de66cb%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-dev+unsubscr...@googlegroups.com.
To post to this group, send email to racket-dev@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/racket-dev/m2d1j3yuj4.wl-stamourv%40eecs.northwestern.edu.
For more options, visit https://groups.google.com/d/optout.

Re: [racket-dev] Inconsistent behavior when doing bitwise arithmetic

[peter . samarin]

Do you have any guess why the resulting numbers vary so much?

On Thursday, October 13, 2016 at 7:20:47 PM UTC+2, mflatt wrote:
>
> Thanks for the report! 
>
> This is a bug in the optimizer's handling of `bitwise-and`, where I 
> made it assume a fixnum result if either argument is a fixnum. That's 
> obviously not correct if the fixnum is negative. I'll push a repair. 
>
> The difference you see when running in a module is because the 
> optimizer constant-folds the calculation, so there's no `bitwise-and` 
> call left to make incorrect assumptions about. 
>
> At Thu, 13 Oct 2016 09:17:52 -0700 (PDT), peter@gmail.com 
>  wrote: 
> > Hi all, 
> > 
> > I get a weird behavior when using bitwise-ior and bitwise-and with large 
> > numbers. Tested on 2 machines (racket 6.6, Ubuntu 16.04 and 14.04): 
> > 
> > Here is the test example: 
> > 
> > #lang racket 
> > (define num #x) ;; remove one f, and the results are 
> fine 
> > in both cases 
> > (for ([i 5]) 
> >   (printf "~a~n" (bitwise-ior (bitwise-and num -2) 0))) 
> > 
> > 
> > When run from DrRacket using ctr+r, it works properly: 
> > 18446744073709551614 
> > 18446744073709551614 
> > 18446744073709551614 
> > 18446744073709551614 
> > 18446744073709551614 
> > 
> > When pasted into DrRacket's REPL (or run from command line via "racket 
> > random-bug.rkt"), the code does not only produce wrong result, but also 
> > seems to be counting something :) 
> > 70112357026588 
> > 70112357038692 
> > 70112357043588 
> > 70112357048576 
> > 70112357053344 
> > 
> > Removing one "f" from the test number results in correct behavior in 
> both 
> > cases. 
> > Very large numbers are are reduced to the same range (around 54 bits). 
> > Replacing the expression (bitwise-and num -2) by the corresponding 
> number 
> > it computes results in correct behavior. 
> > 
> > Cheers, 
> > Peter 
> > 
> > -- 
> > You received this message because you are subscribed to the Google 
> Groups 
> > "Racket Developers" group. 
> > To unsubscribe from this group and stop receiving emails from it, send 
> an 
> > email to racket-dev+...@googlegroups.com . 
> > To post to this group, send email to racke...@googlegroups.com 
> . 
> > To view this discussion on the web visit 
> > 
> https://groups.google.com/d/msgid/racket-dev/d0310d90-0d88-4902-b430-0069b310d7
>  
> > a3%40googlegroups.com. 
> > For more options, visit https://groups.google.com/d/optout. 
>

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to racket-dev+unsubscr...@googlegroups.com.
To post to this group, send email to racket-dev@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/racket-dev/c4b51560-736a-4a19-9b9f-0b28f4de66cb%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.