GitLab

Detail:
 This changes all the C and assembler includes to be a canoncial Unix format.
 Also match include paths to previous commit for EditIntern/DrawIntern/VerIntern
 Finally, also include some minor type fixes (NULL vs 0)
Admin:
 May be some other paths elsewhere in the source I'm not immediately able to fix.  Will address any issues ASAP, since this is a huge change.

Version 5.54. Not tagged

* Improved stdio to allocate extra FILE objects beyond the static array of 16. Number of open files now only limited by memory and the OS.

* FOPEN_MAX in <stdio.h> increased from 8 to 16 to reflect reality.
* Minor corrections to comments in <stdio.h>.
* Library shutdown tidied for modules. In particular, shutdown of I/O was
  broken for modules which had been entered as applications.

Version 5.53. Tagged as 'RISC_OSLib-5_53'

Version 5.52. Tagged as 'RISC_OSLib-5_52'

          * isprint() now true for non-breaking spaces
<math.h>  * FP_INFINITY corrected to FP_INFINITE
          * fma(), fmal(), llrint(), llrintf(), llrintl(),
            llround(), llroundf(), llroundl(),
            remquo(), remquof(), remquol() added.
          * added inlining for remainder() and fmaf() and various
            float functions.
          * math_errhandling removed pending overhaul - it was misleading.
          * expm1() and log1p() now provide decent (< 1 ulp) accuracy,
            rather than only being marginally better than exp(x)-1 and
            log(1+x).
          * new version of copysign() with standard calling conventions.
          * some C99 functions changed to use infinity/NaNs and exceptions
            for error handling - see comments in header file about the
            changes in progress.
<stdio.h>  * output from %a now correctly rounded
<stdlib.h> * strtof() and strtold() added.
<string.h> * strerror() no longer clears _kernel_last_oserror().

Version 5.49. Tagged as 'RISC_OSLib-5_49'

Detail:
  * No longer uses aasm to build h.swis.
  * (Only) compatible with new C compilers: assumes const static data is
    placed in separate read-only areas, and suitable command-line switch
    is used to ensure that library static data is never placed in a zero-
    init area.
  * Stack extension code now thread-safe.
  * Heap thread-safety code (in clib) now uses SWP instruction test results
    from stack extension code (in kernel) - as a side effect, the exported
    symbol _swp_available now exists, if you need to use it elsewhere.
  * Slightly closer to having top-bit-set heap addresses working.
Admin:
  Tested in a Tungsten build, and with Customer W's test suite.

Version 5.48. Tagged as 'RISC_OSLib-5_48'

Detail:
  * _kernel_irqs_disabled() is included in ansilib build, now needed for
    ARM 2 SWP replacement in c.alloc
  * _kernel_entrypoint() for ansilib now initialises relocation offsets in
    root stack chunk, necessary if linking with -zm object code
  * system() implementation now calls DDEUtils_FlushCL if starting a
    short command line - necessary in case previous system() call was a
    long command line directed at a DDE-unaware application
Admin:
  Ansilib changes tested, DDE change not expected to cause problems.

Version 5.47. Not tagged

Retagged as 'RISC_OSLib-5_47'

Detail:
  * Implemented the thread-safety mutex macros for the user mode heap.
  * Created ansilibm, a version of ansilib suitable for building modules.
  * Fixed an overestimation of the size of the workspace passed to
    SharedCLibrary_LibInitModule[APCS_32] by the stubs.
  * Fixed bug in SharedCLibrary_LibInitModule[APCS_32] regarding handling of
    non-reentrant modules - if module code was at a higher address than the
    module data, then all the intervening data would be pointlessly copied
    back over itself with no allowance for volatility.
  * Removed the bsearch source file, since it was only required for APCS-A
    which the library has not supported since the year 2000, and which the
    compiler is now no longer able to build.
Admin:
  ansilibm builds working modules, but thread safety has not been
  extensively tested.

Version 5.47. Tagged as 'RISC_OSLib-5_47'

* _Exit() added.
* Lots of new <math.h> functions (acosh, asinh, atanh, exp2, expm1,
  ilogb, log1p, log2, logb, scalbn, scalbln, cbrt, erf, erfc,
  lgamma, tgamma, nexttoward, fmaf). Float and long double forms
  of every function added; long double forms are included as another
  library object in the stubs rather than the shared library, as they
  just branch to the double form.
* Subnormal/NaN/infinity cases in various <math.h> functions improved.
* Added <tgmath.h>.
* Headers brought into line with CC 5.54.
* RMEnsures added to C library initialisation to try to load minimum
  CallASWI, FPEmulator, CLib. No errors reported if load fails.
* A few pointless inter-file dependencies removed to reduce minimum
  size of included ANSILib.

Version 5.46. Tagged as 'RISC_OSLib-5_46'

Added atoll, strtoll, strtoull, strtoimax, strtoumax, imaxabs and imaxdiv functions to the standard C library.

Version 5.45. Tagged as 'RISC_OSLib-5_45'

PCI added to swis.h
alloc.c updated to handle bigger slots (new code merged from ARM libraries)
Various 32-bit fixes for backtracing, and general trap handling.
Polite "Application is not 32-bit compatible" message.
Headers <stdint.h> and <inttypes.h> fixed to work in non-C99 mode.
txt changed to do new-style Delete behaviour

Version 5.44. Tagged as 'RISC_OSLib-5_44'

* Added two new library chunks, 4 and 5, which contain extensions to the kernel and C library respectively. These have no static data associated with them, just being extensions of the stub tables. The reason for this is to minimise wasted space in programs that don't use the C99 facilities; o.stubs is now a library split into 3 pieces - basic kernel and CLib, extra kernel and extra

  CLib; only the bits a program needs get included.

* Previous extensions to the C library stubs revoked - they now stop at _swix;
  all the new C99 functions now live in chunk 4. Anyone using those new
  functions should relink with new stubs and ensure this C library version.

* printf/scanf now support 64-bit types through "ll" and "j" length modifiers.

* Run-time support for VLAs (__rt_allocauto and __rt_freeauto) added. No
  attempt is currently made to clear up on longjmp or to cope with someone
  changing the kernel allocator while a VLA is active. These would be a
  future enhancement.

* Added complete 64-bit run-time support (48 functions) to kernel library;
  these functions are compatible with the ones used by the ARM ADS. Many of
  the simpler functions will not normally be used by the compiler, as it
  will generate inline code. There is scope for improvement by switching
  in MULL and CLZ-using forms of multiply and divide when possible.

* llabs and lldiv added to C library.

* Header files corrected in a few areas, and changed to match the C compiler.
  <stdint.h> and <stdbool.h> now require the compiler to be in C99 mode
  (as detected using __STDC_VERSION__).


Version 5.41. Tagged as 'RISC_OSLib-5_41'

* Added C99 functions floorf, ceilf, [l]rint[f], trunc[f], [l]round[f], nearbyint[f] and remainder[f].

* Added macros (only) for sinf, cosf, atanf.
* Requires cc 5.43

Version 5.39. Tagged as 'RISC_OSLib-5_39'

* Added C99 extensions to strftime(). All basic specifiers supported, but "E" and "O" modifiers ignored.

* strftime specifiers %c, %X and %x for "C" locale changed to match C99.
* Optimised hypot[f].
* Missing hypotf() declaration added to <math.h>.
* Fixed Shared C Library - inadvertent change to static data size meant it was
  incompatible with the stubs.

Version 5.38. Tagged as 'RISC_OSLib-5_38'

  * snprintf(), vsnprintf(), vfscanf(), vscanf(), vsscanf()
  * hh, j, z and t printf length modifiers (indicating char, intmax_t, size_t
    and ptrdiff_t respectively)
  * Hexadecimal floating-point printing and scanning (%a/%A)
  * %F printf() specifier (upper-case form of %F)
  * Input/output of NaNs and Infinities
  * imaxdiv_t, imaxdiv(), strtoimax(), strtoumax() (simulated through macros)
  * <fenv.h>: feclearexcept(), fegetexceptflag(), feraiseexcept(),
              fesetexceptflag(), fetestexcept(), fegetround(), fesetround(),
              fegetenv(), feholdexcept(), fesetenv(), feupdateenv()
  * FLT_EVAL_METHOD, DECIMAL_DIG
  * hypot(), hypotf(), fabsf(), fdim(), fdimf(), fmax(), fmaxf(), fmin(),
    fminf()
  * INFINITY, NAN, fpclassify(), isfinite(), isinf(), isnan(), isnormal(),
    signbit(), copysign(), copysignf(), nan(), nanf(), nextafter(),
    nextafterf() isgreater(), isgreaterequal(), isless(), islessequal(),
    islessgreater(), isunordered()

This involves adding 36 new entries to the stubs. Current versions of the C
library will not fault client programs with such larger stubs, but will fill
in the extra entries with junk. Programs requiring the new functions will have
to RMEnsure this version of the Shared C Library.

This version of the C library has been fixed so that in future, any extra
unknown stubs entries will be filled in with a branch to zero, rather than
corrupted.

Requires cc 5.41 or later, both to build, and to make use of some of the extra
facilities.


Version 5.37. Tagged as 'RISC_OSLib-5_37'

  Aborts now give standard error messages (Abort on Data Transfer at... etc)
  *ShowRegs now filled in after aborts
  assert(), abort(), "free failed" and standard signal handlers now use Wimp
    error boxes if in the desktop
  Postmortem button on error boxes to view the postmortem

Also, x$multiply, x$divide, __rt_sdiv, x$remainder, x$udivide, __rt_udiv and
x$uremainder optimised.

Version 5.35. Tagged as 'RISC_OSLib-5_35'

* Fixed C library tty so it can output '\b' (backspace).
* Made offsetof() and va_arg() macros work with C++.

Version 5.34. Tagged as 'RISC_OSLib-5_34'

Detail:
  strxfrm was passing the buffers to the Territory module SWI the wrong
    way around so Territory_TransformString copied the destination over
    the source rather than vice versa!
Admin:
  Reported by <mailto:ian.foster@gps.tandb.co.uk>
  Fixes Bugzilla bug #4375


Version 5.33. Tagged as 'RISC_OSLib-5_33'

  Initialised so_far in stdio.c to remove compiler warning.
Admin:
  Built.  strcmptest() still indicates that strcmp() is working OK!

Version 5.32. Tagged as 'RISC_OSLib-5_32'

Detail:
  This alternate implementation returned non-zero for equal strings,
    because the tidyup routine failed to stop examining the differing
    words when it found matching zero bytes.
Admin:
  Tested in softload C library, and in Lazarus ROM image.  strcmp()
    appears to be working per spec now.

Version 5.31. Tagged as 'RISC_OSLib-5_31'

  Fix for RISC_OSLib res() function.
Detail:
  Version 4.89 featured an improved strcmp() routine that didn't actually
    work properly (bug #4085).  The return values had the wrong sign if this
    optimised code was used and the string contained top-bit set chars.
  Removed that rlib.s.asmdefs hack as the comment in the makefile suggested.
    Nothing needs it any more as the assembler sources that require it now
    get via a -PD argument.
  res_ prefix function was still wrong as it failed to detect valid but wrong
    filenames, hence the "Make$Path not defined" errors in the latest tools
    release.
Admin:
  Fixes Bugzilla bug #4085
  Should fix BaseSW's build problems too.

Version 5.29. Tagged as 'RISC_OSLib-5_29'

  Another getenv() bug fixed.
Detail:
  stdio.c contains a vastly improved implementation of fwrite.
  hostsys.h declares _terminate_getenv to remove build warnings.
  armsys.c contains fix to getenv() to stop Omni dying.
Admin:
  Tested on desktop machine for over a week without incident, including
    several heavy fwrite users (WebServe, C compiler)
  fwrite change is documented in Doc/fwrite
  getenv() bug is Bugzilla bug #28

Version 5.27. Tagged as 'RISC_OSLib-5_27'

  OS_ReadVarVal.

Detail:
  getenv calls OS_ReadVarVal to read a system variable into a buffer.
  If the buffer is not big enough it increases the size of the buffer by
  256 bytes and tries again.
  However, the first time it was called, it would have no buffer and
  would call OS_ReadVarVal with r2=-1. The PRM says this "may" return an error
  which should be ignored, but if it did not return an error then getenv
  would not allocate a buffer and the call would fail.
  This change allocates a 256 byte buffer before OS_ReadVarVal is called.

Admin:
  Untested.

Version 5.20. Tagged as 'RISC_OSLib-5_20'

Version 5.17. Tagged as 'RISC_OSLib-5_17'

  Fixed bug in getenv() relating to macro and numeric system variables.
Detail:
  Uses SYMDEFS file to avoid messing around with generating little text
    files, generating assembler source, assembling it etc.  abssym et al
    are construct as symbol definition template files and then updated
    directly by the linker ready for export.  The linker can use these
    files directly without needing them to be assembled into AOF format.
  Fixed getenv so that it can read macro and numeric variables.  With these
    variable types, R2 does not equal -length of the variable value, so
    trying to obtain a variable with a larger value than any previously
    read variable would fail.  This is now fixed.
  This module will refuse to build without link 5.19 or later.
Admin:
  Requires Library 0.54 or later.

  This component will not build with Library 0.53 or earlier.

Version 5.16. Tagged as 'RISC_OSLib-5_16'

* Correction to static base location in modulewrap - was affecting some
  DSL 4000 video software.

Version 5.15. Tagged as 'RISC_OSLib-5_15'

Detail:
  %#x shows a leading 0x (and %#X leads with 0X) only when the value
    to be displayed is NOT zero.
Admin:
  Cross-verified against: C89 standard; C9X standard; gcc 2.95 implementation.
  Built.

Version 5.14. Tagged as 'RISC_OSLib-5_14'

Version 5.06. Tagged as 'RISC_OSLib-5_06'

This is the standard RISC OS convention (cf OS_NewLine) - some apps assume
it when redirecting output.

Version 5.02. Tagged as 'RISC_OSLib-5_02'

Detail:
  Somebody had modified the SWI call to Territory_Collate to be done
    via _swix instead of _kernel_swi, but didn't remove the _kernel_swi_regs
    variable (r) and didn't change the function return value to return the
    new "ret" variable instead of r.r[0] which gives arbitrary results - in
    particular it confuses qsort.
Admin:
  Tested that it builds.

Version 4.97. Tagged as 'RISC_OSLib-4_97'

Detail:
  If qsort is passed an invalid base pointer and the number of entries
    is zero, then due to some pointer artihmetic going wrong, it aborts.
    The fix is to trap zero length arrays and exit quickly.
Admin:
  Required by cmhg prior to version 5.20 which had an explicit fix to
    prevent triggering this bug.
  Reportedly this fix is required by PhotoDesk prior to version 3.03.

Version 4.96. Tagged as 'RISC_OSLib-4_96'

StrongARMfudge flag removed.
StrongARM compatibility set according to AMBKernel flag in Hdr:RISCOS
strftime %U and %W specifiers fixed for week 0.

Version 4.89. Tagged as 'RISC_OSLib-4_89'

Merged Spinner branch changes onto trunk (new RCA messages files and Makefile changes). Convert to srccommit while we're at it.

Removed a few $Revision's to prevent future spurious merge clashes.

Version 4.87. Tagged as 'RISC_OSLib-4_87'

If a module written in C uses function getenv, buffer space
is allocated in the RMA. This buffer was never freed, leading
to RMA leakage when the module was 'RMKill'ed or 'RMReInit'ed.

Changes

 c.armsys
   added function _terminate_getenv to free the buffer

 c.stdlib
   added flag getenv_finalised to the exit_s structure

   initialised getenv_finalised in function _exit_init

   atexit tests getenv_finalised, calls _terminate_getenv

 s.version
   version string changed to 4.86 (27 Jul 1998)