vbcc 0.9g released - Software News - News

Topic

Replies

Last Post

A1222 support in the SDK and problems [1][2]

25 (980)

salass00

Today 17:49

2024 - April - Space Cadet Pinball [1][2]

27 (699)

nbache

Today 11:45

SAM Flex Stuttering Issue workbench or software use / games [1][2][3]

42 (1441)

joerg

Yesterday 17:55

Attempting to upgrade Sam 440 with an R7 240 or HD 7770

4 (192)

Hypex

Yesterday 16:34

Raylib v4.5.0 [1][2][3]

53 (4176)

MartinW

Yesterday 12:34

ACube/AEon - more info about PCIe cards in SAM440, please [1][2][3][4][5]

87 (25845)

geennaam

4/16 15:50

PiStorm - can PPC for classic Amiga make a come back? [1][2]

22 (2507)

geennaam

4/16 15:17

Heretic II X5000 + RX580 bad fps

3 (190)

TheMagicSN

4/16 13:39

Trying to get a Radeon HD 7750 working in an AmigaOne XE [1][2][3][4]

64 (38410)

geennaam

4/16 9:42

MPlayer 1.5 released! [1][2][3][4]

78 (6057)

Maijestro

4/16 9:23

Pegasos2 with RadeonHD/RX via bridge [1][2][3][4][5][6]

110 (10467)

sailor

4/16 7:43

Porting to AmigaOS4 thread [1][2] ... [19][20][21][22]

430 (135847)

Raziel

4/15 20:55

Ragemem source code

3 (234)

sailor

4/15 19:37

Qemu Pegasos II interrupts issue [1][2][3][4][5]

90 (3074)

Maijestro

4/15 18:19

My Amiga Projects [1][2] ... [4][5][6][7]

121 (17593)

kishigo

4/14 22:11

[

Free for all] [

Too Hot To Handle] [

Classifieds] [View all Topics] [View all Forums] [Forum Settings]

Software News : vbcc 0.9g released

Posted by Capehill on 2019/10/8 21:18:07 (969 reads) News by the same author

There is a new vbcc ("vbcc is a highly optimizing portable and retargetable ISO C compiler") release for AmigaOS 2/3/4, MorphOS, PowerUP, WarpOS and Atari MiNT available.

Quote:

04-Oct-2019: vbcc 0.9g release.
Changes since last release:

Implemented hash table with a default size of 1000. Can be controlled by -hash-size=.
Fixed bug with recursive struct definitions.
Fixed bug with C99 designated initializers.
Fixed C99 initializations of const objects.
Fixed bug with mixed dynamic initializations.
Fixed bug with partially initialized structs (C99).
Addded __noinline attribute.
m68k: General improvements of code generation.
m68k: Use ASR-sequence for signed division with a power of two.
m68k: Fixed peephole bug.
m68k: New attribute __fp0ret, to force float-return in fp0, disregarding a -no-fp-return option.
m68k: Fixed bug with unsigned short array indexes > 32767.
m68k: Allow 16-bit multiplication.
m68k: Fix for possible register clobbering in PUSH loops going up.
m68k: Avoid using the same reqister for Dr:Dq in DIV?L.L instructions.
m68k: Fixed pointer differences in 16-bit int backend (vbccm68ks).
m68k: Fixed bug with shiftisdiv on big-endian hosts.
PPC: Optimized memcmp() in vclib and as assembler-inline.
vclib: Fixed snprintf() and vsnprintf() for sizes exactly matching the source.
vclib atari: Made gem.h compile without warning in ANSI/C89 mode.
vclib kick13: Support for 16-bit int. New config files.
vclib kick13: Workaround for IEEE single to double precision conversion in mathieeedoubtrans.library.
Fixed some kick13 clib headers.

Home page

The comments are owned by the author. We aren't responsible for their content.

Author	Thread
blmara	Published: 2019/10/12 12:56 Updated: 2019/10/12 12:56
Just popping in Joined: 12/05/2006 From: Vantaa,Finland Comments: 41	Re:vbcc 0.9g released Great, thanks for supporting Amiga platforms! Marko

Capehill	Published: 2019/10/12 18:32 Updated: 2019/10/12 18:32
Just can't stay away Joined: 07/14/2007 From: Lothric Comments: 1670	Re:vbcc 0.9g released Installation was simple and there seems to be some nice documentation included. I haven't used VBCC for many, many years, so I stumbled into a print format issue where %f/%g were not working. Compiling and linking with +newlib made it work.

salass00	Published: 2019/10/23 23:22 Updated: 2019/10/23 23:23
Just can't stay away Joined: 11/30/2006 From: Finland Comments: 1878	Re:vbcc 0.9g released @Capehill With the default vbcc c-lib you need to link with -lm to get floating point support, same as with clib2 (with gcc). Only newlib is different and has floating point enabled always (there is a libm.a but it's empty).