[sage-devel] Re: BROADCAST MESSAGE

[Frédéric Chapoton]

Trac 1.2 comes with various changes, including display of full names and 
internationalisation. You may want to look in your user preferences to 
customize some of these if you prefer.

Le dimanche 4 septembre 2022 à 01:32:35 UTC+2, Kwankyu Lee a écrit :

> Trac is now in version 1.2. Thanks!

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/cd35c3ab-2330-46cf-9bc6-36ecc08a005dn%40googlegroups.com.

[sage-devel] abort() in __gmpz_realloc of symbolic expression with `gmp: overflow in mpz type`

[Georgi Guninski]

sage 9.0 on ubuntu 20.

When raising symbolic expression to large integer power
i get abort() with stack:

gmp: overflow in mpz type

/lib/x86_64-linux-gnu/libc.so.6(abort+0x12b)[0x7f26d1c7a859]
/lib/x86_64-linux-gnu/libgmp.so.10(__gmpz_realloc+0xad)[0x7f26d048edcd]
/lib/x86_64-linux-gnu/libgmp.so.10(__gmpz_n_pow_ui+0x7e7)[0x7f26d048ba97]
/lib/x86_64-linux-gnu/libpynac.so.18(_ZNK5GiNaC7numeric5powerEl+0x114)[0x7f26683b1c14]

The non-minimal testcase
===
def binnk3u(n,k):  return ( (n/k)**(k))
n1=(2*10**3);d0=29004853178239;n0=SR(log(n1));tt=binnk3u(n0+d0-1,(d0))
print("passed :(")
===

on sagecell i get no output, just an animated rotating circle.

if possible please don't regress the following, which works:

sage: log(pi)**(2**80)
log(pi)^1208925819614629174706176

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAGUWgD_Un568rw_RfAUmS3YaGiwV4sGh49oVsUkGEC3gvbV1rA%40mail.gmail.com.

Re: [sage-devel] abort() in __gmpz_realloc of symbolic expression with `gmp: overflow in mpz type`

[Vincent Delecroix]

Indeed, there is something wrong with the error handling of exponents

{{{
sage: a = 2 ** (2**63)
Traceback (most recent call last):
...
.../sage/rings/integer.pyx in sage.rings.integer.Integer._pow_
   2229 r = smallInteger(1)
   2230 else:
-> 2231 raise OverflowError(f"exponent must be at most {LONG_MAX}")
   2232 if mpz_sgn(exp) >= 0:
   2233 return r

OverflowError: exponent must be at most 9223372036854775807
}}}
but
{{{
sage: a = 2 ** (2**62)
Traceback (most recent call last):
...
.../sage/rings/integer.pyx in sage.rings.integer.Integer.__pow__
   2149
   2150 if type(left) is type(right):
-> 2151 return (left)._pow_(right)
   2152 elif isinstance(left, Element):
   2153 return coercion_model.bin_op(left, right, operator.pow)

.../sage/rings/integer.pyx in sage.rings.integer.Integer._pow_
   2213
   2214 if mpz_fits_slong_p(exp):
-> 2215 return self._pow_long(mpz_get_si(exp))
   2216
   2217 # Raising to an exponent which doesn't fit in a long overflows

.../sage/rings/integer.pyx in sage.rings.integer.Integer._pow_long
   2245 if n > 0:
   2246 x = PY_NEW(Integer)
-> 2247 sig_on()
   2248 mpz_pow_ui(x.value, self.value, n)
   2249 sig_off()

RuntimeError: Aborted
}}}

On Sun, 4 Sept 2022 at 11:58, Georgi Guninski  wrote:
>
> sage 9.0 on ubuntu 20.
>
> When raising symbolic expression to large integer power
> i get abort() with stack:
>
> gmp: overflow in mpz type
>
> /lib/x86_64-linux-gnu/libc.so.6(abort+0x12b)[0x7f26d1c7a859]
> /lib/x86_64-linux-gnu/libgmp.so.10(__gmpz_realloc+0xad)[0x7f26d048edcd]
> /lib/x86_64-linux-gnu/libgmp.so.10(__gmpz_n_pow_ui+0x7e7)[0x7f26d048ba97]
> /lib/x86_64-linux-gnu/libpynac.so.18(_ZNK5GiNaC7numeric5powerEl+0x114)[0x7f26683b1c14]
>
> The non-minimal testcase
> ===
> def binnk3u(n,k):  return ( (n/k)**(k))
> n1=(2*10**3);d0=29004853178239;n0=SR(log(n1));tt=binnk3u(n0+d0-1,(d0))
> print("passed :(")
> ===
>
> on sagecell i get no output, just an animated rotating circle.
>
> if possible please don't regress the following, which works:
>
> sage: log(pi)**(2**80)
> log(pi)^1208925819614629174706176
>
> --
> You received this message because you are subscribed to the Google Groups 
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to sage-devel+unsubscr...@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-devel/CAGUWgD_Un568rw_RfAUmS3YaGiwV4sGh49oVsUkGEC3gvbV1rA%40mail.gmail.com.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAGEwAAkH6kzYXKibDmY23Y6ZhKCQ3UPt2Rwo6iHSvqvttz77qA%40mail.gmail.com.

Re: [sage-devel] abort() in __gmpz_realloc of symbolic expression with `gmp: overflow in mpz type`

[dmo...@deductivepress.ca]

I created trac ticket #34492  to 
continue this discussion, but it may require an upstream fix in gmp.

On Sunday, September 4, 2022 at 7:32:27 AM UTC-6 vdelecroix wrote:

> Indeed, there is something wrong with the error handling of exponents
>
> {{{
> sage: a = 2 ** (2**63)
> Traceback (most recent call last):
> ...
> .../sage/rings/integer.pyx in sage.rings.integer.Integer._pow_
> 2229 r = smallInteger(1)
> 2230 else:
> -> 2231 raise OverflowError(f"exponent must be at most {LONG_MAX}")
> 2232 if mpz_sgn(exp) >= 0:
> 2233 return r
>
> OverflowError: exponent must be at most 9223372036854775807
> }}}
> but
> {{{
> sage: a = 2 ** (2**62)
> Traceback (most recent call last):
> ...
> .../sage/rings/integer.pyx in sage.rings.integer.Integer.__pow__
> 2149
> 2150 if type(left) is type(right):
> -> 2151 return (left)._pow_(right)
> 2152 elif isinstance(left, Element):
> 2153 return coercion_model.bin_op(left, right, operator.pow)
>
> .../sage/rings/integer.pyx in sage.rings.integer.Integer._pow_
> 2213
> 2214 if mpz_fits_slong_p(exp):
> -> 2215 return self._pow_long(mpz_get_si(exp))
> 2216
> 2217 # Raising to an exponent which doesn't fit in a long overflows
>
> .../sage/rings/integer.pyx in sage.rings.integer.Integer._pow_long
> 2245 if n > 0:
> 2246 x = PY_NEW(Integer)
> -> 2247 sig_on()
> 2248 mpz_pow_ui(x.value, self.value, n)
> 2249 sig_off()
>
> RuntimeError: Aborted
> }}}
>
> On Sun, 4 Sept 2022 at 11:58, Georgi Guninski  wrote:
> >
> > sage 9.0 on ubuntu 20.
> >
> > When raising symbolic expression to large integer power
> > i get abort() with stack:
> >
> > gmp: overflow in mpz type
> >
> > /lib/x86_64-linux-gnu/libc.so.6(abort+0x12b)[0x7f26d1c7a859]
> > /lib/x86_64-linux-gnu/libgmp.so.10(__gmpz_realloc+0xad)[0x7f26d048edcd]
> > /lib/x86_64-linux-gnu/libgmp.so.10(__gmpz_n_pow_ui+0x7e7)[0x7f26d048ba97]
> > 
> /lib/x86_64-linux-gnu/libpynac.so.18(_ZNK5GiNaC7numeric5powerEl+0x114)[0x7f26683b1c14]
> >
> > The non-minimal testcase
> > ===
> > def binnk3u(n,k): return ( (n/k)**(k))
> > n1=(2*10**3);d0=29004853178239;n0=SR(log(n1));tt=binnk3u(n0+d0-1,(d0))
> > print("passed :(")
> > ===
> >
> > on sagecell i get no output, just an animated rotating circle.
> >
> > if possible please don't regress the following, which works:
> >
> > sage: log(pi)**(2**80)
> > log(pi)^1208925819614629174706176
> >
> > --
> > You received this message because you are subscribed to the Google 
> Groups "sage-devel" group.
> > To unsubscribe from this group and stop receiving emails from it, send 
> an email to sage-devel+...@googlegroups.com.
> > To view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-devel/CAGUWgD_Un568rw_RfAUmS3YaGiwV4sGh49oVsUkGEC3gvbV1rA%40mail.gmail.com
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/2321c88c-124d-4a86-8fea-ae0a0870973dn%40googlegroups.com.

[sage-devel] Re: BROADCAST MESSAGE

[Eric Gourgoulhon]

Thank you so much Frédéric at al. for all the work in setting up the new 
Trac!

Eric.

Le dimanche 4 septembre 2022 à 09:32:42 UTC+2, Frédéric Chapoton a écrit :

> Trac 1.2 comes with various changes, including display of full names and 
> internationalisation. You may want to look in your user preferences to 
> customize some of these if you prefer.
>
> Le dimanche 4 septembre 2022 à 01:32:35 UTC+2, Kwankyu Lee a écrit :
>
>> Trac is now in version 1.2. Thanks!
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/fe54559d-4a26-4243-a7c8-2b97fe877d2dn%40googlegroups.com.