From 02a5ae97d4604f50121f743ae42ab244a9ea49c1 Mon Sep 17 00:00:00 2001
From: Matthijs Kooijman <matthijs@stdin.nl>
Date: Wed, 12 Feb 2014 19:37:09 +0100
Subject: In HardwareSerial, don't use int for buffer indices
The index attributes have been uint8_t for a while, so there is no point
in using int for local variables. This should allow the compiler to
generate slightly more efficient code, but (at least on gcc 4.8.2) it
also confuses the register allocator, causing this change to increase
code size by 2 bytes instead due to extra push/pop instructions (but
this will probably change in the future if the compiler improves).
---
cores/arduino/HardwareSerial.cpp | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
(limited to 'cores')
diff --git a/cores/arduino/HardwareSerial.cpp b/cores/arduino/HardwareSerial.cpp
index 81b406a..a270ace 100644
--- a/cores/arduino/HardwareSerial.cpp
+++ b/cores/arduino/HardwareSerial.cpp
@@ -171,7 +171,7 @@ int HardwareSerial::read(void)
return -1;
} else {
unsigned char c = _rx_buffer[_rx_buffer_tail];
- _rx_buffer_tail = (unsigned int)(_rx_buffer_tail + 1) % SERIAL_BUFFER_SIZE;
+ _rx_buffer_tail = (uint8_t)(_rx_buffer_tail + 1) % SERIAL_BUFFER_SIZE;
return c;
}
}
@@ -207,7 +207,7 @@ size_t HardwareSerial::write(uint8_t c)
sbi(*_ucsra, TXC0);
return 1;
}
- int i = (_tx_buffer_head + 1) % SERIAL_BUFFER_SIZE;
+ uint8_t i = (_tx_buffer_head + 1) % SERIAL_BUFFER_SIZE;
// If the output buffer is full, there's nothing for it other than to
// wait for the interrupt handler to empty it a bit
--
cgit v1.2.3-18-g5258