Re: [PATCH] lang/python27 -- Fix namespace collision
On 26/02/2019 12:35 am, dimpase+free...@gmail.com wrote:
On 6/20/17 Kubilay Kosak wrote:
On 6/19/17 4:31 AM, Steve Kargl wrote:
On Sun, Jun 18, 2017 at 11:29:05AM -0700, Steve Kargl wrote:
Both IEEE-754 2008 and ISO/IEC TS 18661-4 define the half-cycle
trignometric functions cospi, sinpi, and tanpi. When libm (aka
math.h) grows support for sinpi(x), lang/python27 has a namespace
collision. The attached patch fixes the problem.
Is this issue relevant only for particular (and/or future) FreeBSD
versions ('where libm grows supports for x, y') or independent of base
entirely?
Also, could you open an upstream issue regarding this please, as a
long-term target for all local (Python port) patches is that they are
included upstream.
I've created a Python PR to fix this:
https://github.com/python/cpython/pull/12027
Dima
This also ensures we can document all patches with their relevant
upstream issue/commit references for our future selves and others.
./koobs
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Thank you Dima! :)
I've added the upstream PR to #232792 [1] and will track upstream progress.
[1] https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=232792
./koobs
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: [PATCH] lang/python27 -- Fix namespace collision
On 6/20/17 Kubilay Kosak wrote:
> On 6/19/17 4:31 AM, Steve Kargl wrote:
> On Sun, Jun 18, 2017 at 11:29:05AM -0700, Steve Kargl wrote:
>> Both IEEE-754 2008 and ISO/IEC TS 18661-4 define the half-cycle
>> trignometric functions cospi, sinpi, and tanpi. When libm (aka
>> math.h) grows support for sinpi(x), lang/python27 has a namespace
>> collision. The attached patch fixes the problem.
>>
> Is this issue relevant only for particular (and/or future) FreeBSD
> versions ('where libm grows supports for x, y') or independent of base
> entirely?
> Also, could you open an upstream issue regarding this please, as a
> long-term target for all local (Python port) patches is that they are
> included upstream.
I've created a Python PR to fix this:
https://github.com/python/cpython/pull/12027
Dima
> This also ensures we can document all patches with their relevant
> upstream issue/commit references for our future selves and others.
> ./koobs
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: [PATCH] lang/python27 -- Fix namespace collision
Le 01/08/2017 à 00:36, Steve Kargl a écrit : > Any chance that the patch will be added to lang/python27? > Is there a PR so that it does not get lost ? -- Mathieu Arnold signature.asc Description: OpenPGP digital signature
Re: [PATCH] lang/python27 -- Fix namespace collision
Any chance that the patch will be added to lang/python27?
--
steve
On Sun, Jun 18, 2017 at 11:31:12AM -0700, Steve Kargl wrote:
> On Sun, Jun 18, 2017 at 11:29:05AM -0700, Steve Kargl wrote:
> > Both IEEE-754 2008 and ISO/IEC TS 18661-4 define the half-cycle
> > trignometric functions cospi, sinpi, and tanpi. When libm (aka
> > math.h) grows support for sinpi(x), lang/python27 has a namespace
> > collision. The attached patch fixes the problem.
> >
>
> Well, that's inconvenient. Seems attachments are stripped.
>
> --- Modules/mathmodule.c.orig 2017-06-18 11:09:05.938222000 -0700
> +++ Modules/mathmodule.c 2017-06-18 11:09:56.248307000 -0700
> @@ -71,7 +71,7 @@
> static const double sqrtpi = 1.772453850905516027298167483341145182798;
>
> static double
> -sinpi(double x)
> +my_sinpi(double x)
> {
> double y, r;
> int n;
> @@ -270,7 +270,7 @@
> integer. */
> if (absx > 200.0) {
> if (x < 0.0) {
> -return 0.0/sinpi(x);
> +return 0.0/my_sinpi(x);
> }
> else {
> errno = ERANGE;
> @@ -294,7 +294,7 @@
> }
> z = z * lanczos_g / y;
> if (x < 0.0) {
> -r = -pi / sinpi(absx) / absx * exp(y) / lanczos_sum(absx);
> +r = -pi / my_sinpi(absx) / absx * exp(y) / lanczos_sum(absx);
> r -= z * r;
> if (absx < 140.0) {
> r /= pow(y, absx - 0.5);
> @@ -366,7 +366,7 @@
> (x-0.5)*(log(x+lanczos_g-0.5)-1);
> }
> else {
> -r = log(pi) - log(fabs(sinpi(absx))) - log(absx) -
> +r = log(pi) - log(fabs(my_sinpi(absx))) - log(absx) -
> (log(lanczos_sum(absx)) - lanczos_g +
> (absx-0.5)*(log(absx+lanczos_g-0.5)-1));
> }
>
> --
> Steve
> 20170425 https://www.youtube.com/watch?v=VWUpyCsUKR4
> 20161221 https://www.youtube.com/watch?v=IbCHE-hONow
> ___
> freebsd-ports@freebsd.org mailing list
> https://lists.freebsd.org/mailman/listinfo/freebsd-ports
> To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
--
Steve
20170425 https://www.youtube.com/watch?v=VWUpyCsUKR4
20161221 https://www.youtube.com/watch?v=IbCHE-hONow
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: [PATCH] lang/python27 -- Fix namespace collision
On Tue, Jun 20, 2017 at 01:06:28PM +1000, Kubilay Kocak wrote:
> On 6/19/17 4:31 AM, Steve Kargl wrote:
> > On Sun, Jun 18, 2017 at 11:29:05AM -0700, Steve Kargl wrote:
> >> Both IEEE-754 2008 and ISO/IEC TS 18661-4 define the half-cycle
> >> trignometric functions cospi, sinpi, and tanpi. When libm (aka
> >> math.h) grows support for sinpi(x), lang/python27 has a namespace
> >> collision. The attached patch fixes the problem.
> >>
>
> Is this issue relevant only for particular (and/or future) FreeBSD
> versions ('where libm grows supports for x, y') or independent of base
> entirely?
It will be an issue with a future BSD. It seems to take a long time
before any of my libm improvements are committed. At the moment, bde
and I are probably the only people with sinpi(x) in our libm.
I do, however, anticipate that glibc will get these functions in
the future as portions of ISO/IEC TS 18661-4 have already been
implemented.
> Also, could you open an upstream issue regarding this please, as a
> long-term target for all local (Python port) patches is that they are
> included upstream.
In the 5 minutes that I've searched, I've found https://bugs.python.org/
but there are no instructions on how to submit a bug report.
--
Steve
20170425 https://www.youtube.com/watch?v=VWUpyCsUKR4
20161221 https://www.youtube.com/watch?v=IbCHE-hONow
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: [PATCH] lang/python27 -- Fix namespace collision
On 6/19/17 4:31 AM, Steve Kargl wrote:
> On Sun, Jun 18, 2017 at 11:29:05AM -0700, Steve Kargl wrote:
>> Both IEEE-754 2008 and ISO/IEC TS 18661-4 define the half-cycle
>> trignometric functions cospi, sinpi, and tanpi. When libm (aka
>> math.h) grows support for sinpi(x), lang/python27 has a namespace
>> collision. The attached patch fixes the problem.
>>
Hi Steve,
Is this issue relevant only for particular (and/or future) FreeBSD
versions ('where libm grows supports for x, y') or independent of base
entirely?
Also, could you open an upstream issue regarding this please, as a
long-term target for all local (Python port) patches is that they are
included upstream.
This also ensures we can document all patches with their relevant
upstream issue/commit references for our future selves and others.
./koobs
> Well, that's inconvenient. Seems attachments are stripped.
>
> --- Modules/mathmodule.c.orig 2017-06-18 11:09:05.938222000 -0700
> +++ Modules/mathmodule.c 2017-06-18 11:09:56.248307000 -0700
> @@ -71,7 +71,7 @@
> static const double sqrtpi = 1.772453850905516027298167483341145182798;
>
> static double
> -sinpi(double x)
> +my_sinpi(double x)
> {
> double y, r;
> int n;
> @@ -270,7 +270,7 @@
> integer. */
> if (absx > 200.0) {
> if (x < 0.0) {
> -return 0.0/sinpi(x);
> +return 0.0/my_sinpi(x);
> }
> else {
> errno = ERANGE;
> @@ -294,7 +294,7 @@
> }
> z = z * lanczos_g / y;
> if (x < 0.0) {
> -r = -pi / sinpi(absx) / absx * exp(y) / lanczos_sum(absx);
> +r = -pi / my_sinpi(absx) / absx * exp(y) / lanczos_sum(absx);
> r -= z * r;
> if (absx < 140.0) {
> r /= pow(y, absx - 0.5);
> @@ -366,7 +366,7 @@
> (x-0.5)*(log(x+lanczos_g-0.5)-1);
> }
> else {
> -r = log(pi) - log(fabs(sinpi(absx))) - log(absx) -
> +r = log(pi) - log(fabs(my_sinpi(absx))) - log(absx) -
> (log(lanczos_sum(absx)) - lanczos_g +
> (absx-0.5)*(log(absx+lanczos_g-0.5)-1));
> }
>
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
Re: [PATCH] lang/python27 -- Fix namespace collision
On Sun, Jun 18, 2017 at 11:29:05AM -0700, Steve Kargl wrote:
> Both IEEE-754 2008 and ISO/IEC TS 18661-4 define the half-cycle
> trignometric functions cospi, sinpi, and tanpi. When libm (aka
> math.h) grows support for sinpi(x), lang/python27 has a namespace
> collision. The attached patch fixes the problem.
>
Well, that's inconvenient. Seems attachments are stripped.
--- Modules/mathmodule.c.orig 2017-06-18 11:09:05.938222000 -0700
+++ Modules/mathmodule.c2017-06-18 11:09:56.248307000 -0700
@@ -71,7 +71,7 @@
static const double sqrtpi = 1.772453850905516027298167483341145182798;
static double
-sinpi(double x)
+my_sinpi(double x)
{
double y, r;
int n;
@@ -270,7 +270,7 @@
integer. */
if (absx > 200.0) {
if (x < 0.0) {
-return 0.0/sinpi(x);
+return 0.0/my_sinpi(x);
}
else {
errno = ERANGE;
@@ -294,7 +294,7 @@
}
z = z * lanczos_g / y;
if (x < 0.0) {
-r = -pi / sinpi(absx) / absx * exp(y) / lanczos_sum(absx);
+r = -pi / my_sinpi(absx) / absx * exp(y) / lanczos_sum(absx);
r -= z * r;
if (absx < 140.0) {
r /= pow(y, absx - 0.5);
@@ -366,7 +366,7 @@
(x-0.5)*(log(x+lanczos_g-0.5)-1);
}
else {
-r = log(pi) - log(fabs(sinpi(absx))) - log(absx) -
+r = log(pi) - log(fabs(my_sinpi(absx))) - log(absx) -
(log(lanczos_sum(absx)) - lanczos_g +
(absx-0.5)*(log(absx+lanczos_g-0.5)-1));
}
--
Steve
20170425 https://www.youtube.com/watch?v=VWUpyCsUKR4
20161221 https://www.youtube.com/watch?v=IbCHE-hONow
___
freebsd-ports@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
[PATCH] lang/python27 -- Fix namespace collision
Both IEEE-754 2008 and ISO/IEC TS 18661-4 define the half-cycle trignometric functions cospi, sinpi, and tanpi. When libm (aka math.h) grows support for sinpi(x), lang/python27 has a namespace collision. The attached patch fixes the problem. -- Steve 20170425 https://www.youtube.com/watch?v=VWUpyCsUKR4 20161221 https://www.youtube.com/watch?v=IbCHE-hONow ___ freebsd-ports@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "freebsd-ports-unsubscr...@freebsd.org"
