- CLN-list - ginac.de

adding two different precision number
by Po-Hsun Tseng 09 Feb '20

09 Feb '20

Dear CLN developers, 1. Does CLN support performing two different precision number with basic overloading operators (+,-,*, and /)? 2. If yes, how does CLN treat that? eg. zero padding on the precision-less number before doing operation? 3. If no, how can i achieve that? thanks. Po-Hsun Tseng (Rocky)

2 2

array support
by Po-Hsun Tseng 09 Feb '20

09 Feb '20

Dear CLN developers, Can I store CLN-like numbers in an array? Po-Hsun Tseng (Rocky)

2 1

Patchset: build with CMake
by Alexey Sheplyakov 02 Jan '20

02 Jan '20

2 1

Patch: fix factorial calculation on 64-bit windows
by Alexey Sheplyakov 24 Dec '19

24 Dec '19

1 0

CLN 1.3.6 has been released
by Richard B. Kreckel 05 Dec '19

05 Dec '19

Hi! This is a minor update fixing cross-compilation (broken in the last release). Get it from <http://www.ginac.de/CLN/>. -richy. -- Richard B. Kreckel <http://in.terlu.de/~kreckel/>

1 0

cln 1.3.5
by Henrik Grimler 05 Dec '19

05 Dec '19

Hi, When I try to cross-compile 1.3.5 for android, with clang 8.0.7 (same with clang 9.0.0), it generates an intparam.h with: #error "Integers of type int have no binary representation!!" #error "Integers of type long have no binary representation!!" #error "Integers of type long long have no binary representation!!"[...] #error "Integer types int and unsigned int have different sizes!!" #error "Integer types long and unsigned long have different sizes!!" #error "Integer types long long and unsigned long long have different sizes!!" #error "Type char * does not fit into an intptr_t!!" I do not get the same errors with 1.3.4. It seems to due to the typedef int[<negative>] -> static_assert change done in 2f7994014542920ad6ca726fbd8d023a5f49abc2. Reverting that commit makes the build succeed again. If I compile the old type of test, everything works fine, so int main() { typedef int verify[2*((int)((int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2) == 0) - 1]; ; return 0; } compiles without errors, while the same test with static_assert int main() { static_assert((int)((int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2 * (int)2) == 0, ""); ; return 0; } fails with "test_new.cpp:3:17: error: static_assert expression is not an integral constant expression" I unfortunately do not know what the proper fix would be, but reverting 2f7994014542920ad6ca726fbd8d023a5f49abc2 makes the build succeed again at least. Best regards, Henrik Grimler

3 2

CLN 1.3.5 has been released
by Richard B. Kreckel 17 Nov '19

17 Nov '19

Hi! This release fixes a bug in scale_float(), which could throw underflow_exception even if cl_inhibit_floating_point_underflow was set to true. On the portability side, support for compilation on Windows has been improved, see the INSTALL.windows file. Also, the Elbrus 2000, Risc-V64, and MIPS rel 6 platforms are now supported. Many compiler warnings have been addressed, too. Download it from the usual place: <http://www.ginac.de/CLN/>. -richy. -- Richard B. Kreckel <http://in.terlu.de/~kreckel/>

1 0

MinGW64 build
by Robert Szalai 05 Nov '19

05 Nov '19

Hi all, I've been trying to build CLN from windows 64, with some success. It compiles fine, but does not work. The problem is that WIN64 is LLP64, and not ILP64 as Linux. CLN uses long to store pointer data, but size_t or ssize_t should be used as it is guaranteed to be the size of the pointer everywhere. The attached patch was my attempt. Could somebody tell me where else should I look? Pointers are also encoding some metadata in the bits that are set to zero by alingnment, what is this data? There is also some hashing of pointers. Thanks, Robert

6 21

Floating point underflow
by Jan Rheinländer 01 Nov '19

01 Nov '19

Hi, trying to inhibit floating point underflow exceptions, but it doesn't seem to work. What am I doing wrong? #include <cln/cln.h> int main() { cln::cl_inhibit_floating_point_underflow = true; cln::cl_N x = 10.0; cln::cl_N ep = 100.0; cln::cl_N y = cln::expt(x, ep); cln::cl_N z = cln::expt(x, -y); } Running this program still gives me an underflow exception. Thanks for any hints. Jan

2 1

CLN switched to C++11
by Richard B. Kreckel 28 Oct '19

28 Oct '19

Hi, I've just just bumped the minimum C++ language support for the compiler to C++11. I did this because C++11 is a better language and it's widely supported by now. Sticking to C++98 would require us to sometimes write convoluted patches. All in all it's a minor thing: configure makes sure that proper compiler options are chosen and the change doesn't even break the ABI. Still, I thought I should announce this here. All my best, -richy.

3 8