I thought that jtmf inspects the header and vectors to your gmp code to free it (line 1485).  jtmf accepts any A block, including A blocks in indirect values, but it does the free only on blocks it allocated.

Going away for 1-2 hours,

hhr

On 11/11/2023 9:28 AM, Raul Miller wrote:
Hmm...

My thinking was that jtmf was the implementation for the mf() macro
which the mfgmp() macro is parallel to.

This corresponds also to the structure and use of frcommon().

So I guess my question to you is: jtmf is capable of safely freeing
gmp allocated memory, what's the point of the frcommon() macro (and
the fr() and frgmp() macros which rely on it)?

Thanks,


--
Raul



On Sat, Nov 11, 2023 at 9:19 AM Henry Rich <[email protected]> wrote:
But it was flagged that way in the block, no?  So it should be be freed
by the gmp allocator.


hhr

On 11/11/2023 9:16 AM, Raul Miller wrote:
The free was memory which had not been allocated by J's memory
allocation process.

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to