Hi All,
I have the following code below, with "default" signed char type.
-----------
#define BIT4 0b00010000
#define P_IO_Ctrl_Base 0x2080
#define P_IO_PortC_Data *(volatile unsigned char *) (P_IO_Ctrl_Base+0x1A)
#define DrvGPIO_SetBit(GPIO,BITIO) GPIO |= BITIO;
#define DrvGPIO_ClrBit(GPIO,BITIO) GPIO &= ~BITIO;
void main()
{
DrvGPIO_ClrBit(P_IO_PortC_Data,BIT4);
}
----------
I add `--fverbose-asm` option to see more details on the codegen. What
I'm curious about is the code snipt below:
----------
ld a,(#0x209A)
; peephole 17 loaded a from (#0x209A) directly instead of using hl.
; genCast
; genAnd
and a, #0xEF
---------
Although the genCast doesn't emit code, I am wondering why we need a
casting here. So I study C integer conversion rules [1] a little bit,
and here is my reasoning.
1. BIT4 => signed char
2. ~BIT4 => int
3. and we now have the following situation,
P_IO_PortC_Data &= ~BIT4
^^^^^^^^^^^^^^^ ^^^^^
unsigned char int
| *
v
int
the reason we have genCast is because we have unsigned char to int
promotion.
Am I right or wrong about the C integer conversion rules? :-)
Thanks in advance.
[1]
https://www.securecoding.cert.org/confluence/display/seccode/INT02-C.+Understand+integer+conversion+rules
Regards,
chenwj
--
Wei-Ren Chen (陳韋任)
Homepage: http://people.cs.nctu.edu.tw/~chenwj
------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk
_______________________________________________
Sdcc-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/sdcc-user
