RE: [Gcl-devel] load-time value support, closed compile bugs in ANSI test suite

[Mike Thomas]

Hi Camm.

| Went through the load-time-value examples in the spec, and all appears
| OK AFAICT.  Please give it a thorough working over.  There is even
| some partial usage of the read-only-p flag to permit
| copying/coalescing in certain cases.  Please let me know if I've made
| any mistakes here.

Can't say for sure yet whether this is caused by your changes but yesterday
I found that a complete rebuild of GCL on Windows now falls over deep in the
garbage collector at this point in the build:

...

Initializing pcl_env.o
Initializing pcl_gazonk8.o

Warning: DESCRIBE is being redefined.Initializing pcl_compat.o
Initializing pcl_precom1.o
Initializing pcl_precom2.o

Unrecoverable error: Segmentation violation..
 This application has requested the Runtime to terminate it in an unusual
way.
Please contact the application's support team for more information. make[1]:
*** [saved_ansi_gcl] Error 3
rm init_ansi_gcl.lsp.tmp
make[1]: Leaving directory `/c/cvs/head/gcl/unixport'
make: *** [unixport/saved_ansi_gcl] Error 2

...

This has gone wrong somewhere between my report on the AMD64 about a week
ago and yesterday morning, my time.  Hope to look more closely at it in the
next week if you aren't seeing the same problem on Unix.  (It happens on
both PIV and AMD64 machines by the way.)  Ironically, I discovered this
while compiling some ANSI test failure fixes for the "get" function - I had
decided not to bother with the compiler return values as I thought it might
be viewed as cheating just to return nil for the secondary and tertiary
values without addressing the condition stuff!!

Cheers

Mike Thomas.