ePenguin
/
ePenguin-Software-Framework
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
stable
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'a2aee475ea'
${ noResults }
ePenguin-Software-Framework/toolchains/avr/share/doc/gcc/AVR-Options.html

514 lines
33 KiB
HTML
Raw Normal View History Unescape Escape

the monkey dances and laughs and sings and dances
4 years ago
<html lang="en">
<head>
<title>AVR Options - Using the GNU Compiler Collection (GCC)</title>
<meta http-equiv="Content-Type" content="text/html">
<meta name="description" content="Using the GNU Compiler Collection (GCC)">
<meta name="generator" content="makeinfo 4.13">
<link title="Top" rel="start" href="index.html#Top">
<link rel="up" href="Submodel-Options.html#Submodel-Options" title="Submodel Options">
<link rel="prev" href="ARM-Options.html#ARM-Options" title="ARM Options">
<link rel="next" href="Blackfin-Options.html#Blackfin-Options" title="Blackfin Options">
<link href="http://www.gnu.org/software/texinfo/" rel="generator-home" title="Texinfo Homepage">
<!--
Copyright (C) 1988-2015 Free Software Foundation, Inc.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.3 or
any later version published by the Free Software Foundation; with the
Invariant Sections being ``Funding Free Software'', the Front-Cover
Texts being (a) (see below), and with the Back-Cover Texts being (b)
(see below). A copy of the license is included in the section entitled
``GNU Free Documentation License''.
(a) The FSF's Front-Cover Text is:
A GNU Manual
(b) The FSF's Back-Cover Text is:
You have freedom to copy and modify this GNU Manual, like GNU
software. Copies published by the Free Software Foundation raise
funds for GNU development.-->
<meta http-equiv="Content-Style-Type" content="text/css">
<style type="text/css"><!--
pre.display { font-family:inherit }
pre.format { font-family:inherit }
pre.smalldisplay { font-family:inherit; font-size:smaller }
pre.smallformat { font-family:inherit; font-size:smaller }
pre.smallexample { font-size:smaller }
pre.smalllisp { font-size:smaller }
span.sc { font-variant:small-caps }
span.roman { font-family:serif; font-weight:normal; }
span.sansserif { font-family:sans-serif; font-weight:normal; }
--></style>
</head>
<body>
<div class="node">
<a name="AVR-Options"></a>
<p>
Next:&nbsp;<a rel="next" accesskey="n" href="Blackfin-Options.html#Blackfin-Options">Blackfin Options</a>,
Previous:&nbsp;<a rel="previous" accesskey="p" href="ARM-Options.html#ARM-Options">ARM Options</a>,
Up:&nbsp;<a rel="up" accesskey="u" href="Submodel-Options.html#Submodel-Options">Submodel Options</a>
<hr>
</div>
<h4 class="subsection">3.17.5 AVR Options</h4>
<p><a name="index-AVR-Options-1432"></a>
These options are defined for AVR implementations:
<dl>
<dt><code>-mmcu=</code><var>mcu</var><dd><a name="index-mmcu-1433"></a>Specify Atmel AVR instruction set architectures (ISA) or MCU type.
<p>The default for this option is&nbsp;&lsquo;<samp><span class="samp">avr2</span></samp>&rsquo;.
<p>GCC supports the following AVR devices and ISAs:
<!-- Copyright (C) 2012-2015 Free Software Foundation, Inc. -->
<!-- This is part of the GCC manual. -->
<!-- For copying conditions, see the file gcc/doc/include/fdl.texi. -->
<!-- This file is generated automatically using -->
<!-- gcc/config/avr/gen-avr-mmcu-texi.c from: -->
<!-- gcc/config/avr/avr-arch.h -->
<!-- gcc/config/avr/avr-devices.c -->
<!-- gcc/config/avr/avr-mcus.def -->
<!-- Please do not edit manually. -->
<dl>
<dt><code>avr2</code><dd>&ldquo;Classic&rdquo; devices with up to 8&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>attiny22</code>, <code>attiny26</code>, <code>at90c8534</code>, <code>at90s2313</code>, <code>at90s2323</code>, <code>at90s2333</code>, <code>at90s2343</code>, <code>at90s4414</code>, <code>at90s4433</code>, <code>at90s4434</code>, <code>at90s8515</code>, <code>at90s8535</code>.
<br><dt><code>avr25</code><dd>&ldquo;Classic&rdquo; devices with up to 8&nbsp;KiB of program memory and with the <code>MOVW</code> instruction.
<br><var>mcu</var>&nbsp;= <code>ata5272</code>, <code>ata6616c</code>, <code>attiny13</code>, <code>attiny13a</code>, <code>attiny2313</code>, <code>attiny2313a</code>, <code>attiny24</code>, <code>attiny24a</code>, <code>attiny25</code>, <code>attiny261</code>, <code>attiny261a</code>, <code>attiny43u</code>, <code>attiny4313</code>, <code>attiny44</code>, <code>attiny44a</code>, <code>attiny441</code>, <code>attiny45</code>, <code>attiny461</code>, <code>attiny461a</code>, <code>attiny48</code>, <code>attiny828</code>, <code>attiny84</code>, <code>attiny84a</code>, <code>attiny841</code>, <code>attiny85</code>, <code>attiny861</code>, <code>attiny861a</code>, <code>attiny87</code>, <code>attiny88</code>, <code>at86rf401</code>.
<br><dt><code>avr3</code><dd>&ldquo;Classic&rdquo; devices with 16&nbsp;KiB up to 64&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>at43usb355</code>, <code>at76c711</code>.
<br><dt><code>avr31</code><dd>&ldquo;Classic&rdquo; devices with 128&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>atmega103</code>, <code>at43usb320</code>.
<br><dt><code>avr35</code><dd>&ldquo;Classic&rdquo; devices with 16&nbsp;KiB up to 64&nbsp;KiB of program memory and with the <code>MOVW</code> instruction.
<br><var>mcu</var>&nbsp;= <code>ata5505</code>, <code>ata6617c</code>, <code>ata664251</code>, <code>atmega16u2</code>, <code>atmega32u2</code>, <code>atmega8u2</code>, <code>attiny1634</code>, <code>attiny167</code>, <code>at90usb162</code>, <code>at90usb82</code>.
<br><dt><code>avr4</code><dd>&ldquo;Enhanced&rdquo; devices with up to 8&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>ata6285</code>, <code>ata6286</code>, <code>ata6289</code>, <code>ata6612c</code>, <code>atmega48</code>, <code>atmega48a</code>, <code>atmega48p</code>, <code>atmega48pa</code>, <code>atmega48pb</code>, <code>atmega8</code>, <code>atmega8a</code>, <code>atmega8hva</code>, <code>atmega8515</code>, <code>atmega8535</code>, <code>atmega88</code>, <code>atmega88a</code>, <code>atmega88p</code>, <code>atmega88pa</code>, <code>atmega88pb</code>, <code>at90pwm1</code>, <code>at90pwm2</code>, <code>at90pwm2b</code>, <code>at90pwm3</code>, <code>at90pwm3b</code>, <code>at90pwm81</code>.
<br><dt><code>avr5</code><dd>&ldquo;Enhanced&rdquo; devices with 16&nbsp;KiB up to 64&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>ata5702m322</code>, <code>ata5782</code>, <code>ata5790</code>, <code>ata5790n</code>, <code>ata5791</code>, <code>ata5795</code>, <code>ata5831</code>, <code>ata6613c</code>, <code>ata6614q</code>, <code>ata8210</code>, <code>ata8510</code>, <code>atmega16</code>, <code>atmega16a</code>, <code>atmega16hva</code>, <code>atmega16hva2</code>, <code>atmega16hvb</code>, <code>atmega16hvbrevb</code>, <code>atmega16m1</code>, <code>atmega16u4</code>, <code>atmega161</code>, <code>atmega162</code>, <code>atmega163</code>, <code>atmega164a</code>, <code>atmega164p</code>, <code>atmega164pa</code>, <code>atmega165</code>, <code>atmega165a</code>, <code>atmega165p</code>, <code>atmega165pa</code>, <code>atmega168</code>, <code>atmega168a</code>, <code>atmega168p</code>, <code>atmega168pa</code>, <code>atmega168pb</code>, <code>atmega169</code>, <code>atmega169a</code>, <code>atmega169p</code>, <code>atmega169pa</code>, <code>atmega32</code>, <code>atmega32a</code>, <code>atmega32c1</code>, <code>atmega32hvb</code>, <code>atmega32hvbrevb</code>, <code>atmega32m1</code>, <code>atmega32u4</code>, <code>atmega32u6</code>, <code>atmega323</code>, <code>atmega324a</code>, <code>atmega324p</code>, <code>atmega324pa</code>, <code>atmega325</code>, <code>atmega325a</code>, <code>atmega325p</code>, <code>atmega325pa</code>, <code>atmega3250</code>, <code>atmega3250a</code>, <code>atmega3250p</code>, <code>atmega3250pa</code>, <code>atmega328</code>, <code>atmega328p</code>, <code>atmega328pb</code>, <code>atmega329</code>, <code>atmega329a</code>, <code>atmega329p</code>, <code>atmega329pa</code>, <code>atmega3290</code>, <code>atmega3290a</code>, <code>atmega3290p</code>, <code>atmega3290pa</code>, <code>atmega406</code>, <code>atmega64</code>, <code>atmega64a</code>, <code>atmega64c1</code>, <code>atmega64hve</code>, <code>atmega64hve2</code>, <code>atmega64m1</code>, <code>atmega64rfr2</code>, <code>atmega640</code>, <code>atmega644</code>, <code>atmega644a</code>, <code>atmega644p</code>, <code>atmega644pa</code>, <code>atmega644rfr2</code>, <code>atmega645</code>, <code>atmega645a</code>, <code>atmega645p</code>, <code>atmega6450</code>, <code>atmega6450a</code>, <code>atmega6450p</code>, <code>atmega649</code>, <code>atmega649a</code>, <code>atmega649p</code>, <code>atmega6490</code>, <code>atmega6490a</code>, <code>atmega6490p</code>, <code>at90can32</code>, <code>at90can64</code>, <code>at90pwm161</code>, <code>at90pwm216</code>, <code>at90pwm316</code>, <code>at90scr100</code>, <code>at90usb646</code>, <code>at90usb647</code>, <code>at94k</code>, <code>m3000</code>.
<br><dt><code>avr51</code><dd>&ldquo;Enhanced&rdquo; devices with 128&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>atmega128</code>, <code>atmega128a</code>, <code>atmega128rfa1</code>, <code>atmega128rfr2</code>, <code>atmega1280</code>, <code>atmega1281</code>, <code>atmega1284</code>, <code>atmega1284p</code>, <code>atmega1284rfr2</code>, <code>at90can128</code>, <code>at90usb1286</code>, <code>at90usb1287</code>.
<br><dt><code>avr6</code><dd>&ldquo;Enhanced&rdquo; devices with 3-byte PC, i.e. with more than 128&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>atmega256rfr2</code>, <code>atmega2560</code>, <code>atmega2561</code>, <code>atmega2564rfr2</code>.
<br><dt><code>avrxmega2</code><dd>&ldquo;XMEGA&rdquo; devices with more than 8&nbsp;KiB and up to 64&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>atxmega16a4</code>, <code>atxmega16a4u</code>, <code>atxmega16c4</code>, <code>atxmega16d4</code>, <code>atxmega16e5</code>, <code>atxmega32a4</code>, <code>atxmega32a4u</code>, <code>atxmega32c3</code>, <code>atxmega32c4</code>, <code>atxmega32d3</code>, <code>atxmega32d4</code>, <code>atxmega32e5</code>, <code>atxmega8e5</code>.
<br><dt><code>avrxmega3</code><dd>&ldquo;XMEGA&rdquo; devices with up to 64&nbsp;KiB of combined program memory and RAM, and with program memory visible in the RAM address space.
<br><var>mcu</var>&nbsp;= <code>attiny1614</code>, <code>attiny1616</code>, <code>attiny1617</code>, <code>attiny212</code>, <code>attiny214</code>, <code>attiny3216</code>, <code>attiny3217</code>, <code>attiny412</code>, <code>attiny414</code>, <code>attiny416</code>, <code>attiny417</code>, <code>attiny814</code>, <code>attiny816</code>, <code>attiny817</code>.
<br><dt><code>avrxmega4</code><dd>&ldquo;XMEGA&rdquo; devices with more than 64&nbsp;KiB and up to 128&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>atxmega64a3</code>, <code>atxmega64a3u</code>, <code>atxmega64a4u</code>, <code>atxmega64b1</code>, <code>atxmega64b3</code>, <code>atxmega64c3</code>, <code>atxmega64d3</code>, <code>atxmega64d4</code>.
<br><dt><code>avrxmega5</code><dd>&ldquo;XMEGA&rdquo; devices with more than 64&nbsp;KiB and up to 128&nbsp;KiB of program memory and more than 64&nbsp;KiB of RAM.
<br><var>mcu</var>&nbsp;= <code>atxmega64a1</code>, <code>atxmega64a1u</code>.
<br><dt><code>avrxmega6</code><dd>&ldquo;XMEGA&rdquo; devices with more than 128&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>atxmega128a3</code>, <code>atxmega128a3u</code>, <code>atxmega128b1</code>, <code>atxmega128b3</code>, <code>atxmega128c3</code>, <code>atxmega128d3</code>, <code>atxmega128d4</code>, <code>atxmega192a3</code>, <code>atxmega192a3u</code>, <code>atxmega192c3</code>, <code>atxmega192d3</code>, <code>atxmega256a3</code>, <code>atxmega256a3b</code>, <code>atxmega256a3bu</code>, <code>atxmega256a3u</code>, <code>atxmega256c3</code>, <code>atxmega256d3</code>, <code>atxmega384c3</code>, <code>atxmega384d3</code>.
<br><dt><code>avrxmega7</code><dd>&ldquo;XMEGA&rdquo; devices with more than 128&nbsp;KiB of program memory and more than 64&nbsp;KiB of RAM.
<br><var>mcu</var>&nbsp;= <code>atxmega128a1</code>, <code>atxmega128a1u</code>, <code>atxmega128a4u</code>.
<br><dt><code>avrtiny</code><dd>&ldquo;TINY&rdquo; Tiny core devices with 512&nbsp;B up to 4&nbsp;KiB of program memory.
<br><var>mcu</var>&nbsp;= <code>attiny10</code>, <code>attiny20</code>, <code>attiny4</code>, <code>attiny40</code>, <code>attiny5</code>, <code>attiny9</code>.
<br><dt><code>avr1</code><dd>This ISA is implemented by the minimal AVR core and supported for assembler only.
<br><var>mcu</var>&nbsp;= <code>attiny11</code>, <code>attiny12</code>, <code>attiny15</code>, <code>attiny28</code>, <code>at90s1200</code>.
</dl>
<br><dt><code>-mabsdata</code><dd><a name="index-mabsdata-1434"></a>
Assume that all data in static storage can be accessed by LDS / STS
instructions. This option has only an effect on reduced Tiny devices like
ATtiny40.
<br><dt><code>-maccumulate-args</code><dd><a name="index-maccumulate_002dargs-1435"></a>Accumulate outgoing function arguments and acquire/release the needed
stack space for outgoing function arguments once in function
prologue/epilogue. Without this option, outgoing arguments are pushed
before calling a function and popped afterwards.
<p>Popping the arguments after the function call can be expensive on
AVR so that accumulating the stack space might lead to smaller
executables because arguments need not to be removed from the
stack after such a function call.
<p>This option can lead to reduced code size for functions that perform
several calls to functions that get their arguments on the stack like
calls to printf-like functions.
<br><dt><code>-mbranch-cost=</code><var>cost</var><dd><a name="index-mbranch_002dcost-1436"></a>Set the branch costs for conditional branch instructions to
<var>cost</var>. Reasonable values for <var>cost</var> are small, non-negative
integers. The default branch cost is 0.
<br><dt><code>-mcall-prologues</code><dd><a name="index-mcall_002dprologues-1437"></a>Functions prologues/epilogues are expanded as calls to appropriate
subroutines. Code size is smaller.
<br><dt><code>-mint8</code><dd><a name="index-mint8-1438"></a>Assume <code>int</code> to be 8-bit integer. This affects the sizes of all types: a
<code>char</code> is 1 byte, an <code>int</code> is 1 byte, a <code>long</code> is 2 bytes,
and <code>long long</code> is 4 bytes. Please note that this option does not
conform to the C standards, but it results in smaller code
size.
<br><dt><code>-mn-flash=</code><var>num</var><dd><a name="index-mn_002dflash-1439"></a>Assume that the flash memory has a size of
<var>num</var> times 64&nbsp;KiB.
<br><dt><code>-mno-interrupts</code><dd><a name="index-mno_002dinterrupts-1440"></a>Generated code is not compatible with hardware interrupts.
Code size is smaller.
<br><dt><code>-mrelax</code><dd><a name="index-mrelax-1441"></a>Try to replace <code>CALL</code> resp. <code>JMP</code> instruction by the shorter
<code>RCALL</code> resp. <code>RJMP</code> instruction if applicable.
Setting <samp><span class="option">-mrelax</span></samp> just adds the <samp><span class="option">--mlink-relax</span></samp> option to
the assembler's command line and the <samp><span class="option">--relax</span></samp> option to the
linker's command line.
<p>Jump relaxing is performed by the linker because jump offsets are not
known before code is located. Therefore, the assembler code generated by the
compiler is the same, but the instructions in the executable may
differ from instructions in the assembler code.
<p>Relaxing must be turned on if linker stubs are needed, see the
section on <code>EIND</code> and linker stubs below.
<br><dt><code>-mrmw</code><dd><a name="index-mrmw-1442"></a>Assume that the device supports the Read-Modify-Write
instructions <code>XCH</code>, <code>LAC</code>, <code>LAS</code> and <code>LAT</code>.
<br><dt><code>-mshort-calls</code><dd><a name="index-mshort_002dcalls-1443"></a>
Assume that <code>RJMP</code> and <code>RCALL</code> can target the whole
program memory.
<p>This option is used internally for multilib selection. It is
not an optimization option, and you don't need to set it by hand.
<br><dt><code>-msp8</code><dd><a name="index-msp8-1444"></a>Treat the stack pointer register as an 8-bit register,
i.e. assume the high byte of the stack pointer is zero.
In general, you don't need to set this option by hand.
<p>This option is used internally by the compiler to select and
build multilibs for architectures <code>avr2</code> and <code>avr25</code>.
These architectures mix devices with and without <code>SPH</code>.
For any setting other than <samp><span class="option">-mmcu=avr2</span></samp> or <samp><span class="option">-mmcu=avr25</span></samp>
the compiler driver adds or removes this option from the compiler
proper's command line, because the compiler then knows if the device
or architecture has an 8-bit stack pointer and thus no <code>SPH</code>
register or not.
<br><dt><code>-mstrict-X</code><dd><a name="index-mstrict_002dX-1445"></a>Use address register <code>X</code> in a way proposed by the hardware. This means
that <code>X</code> is only used in indirect, post-increment or
pre-decrement addressing.
<p>Without this option, the <code>X</code> register may be used in the same way
as <code>Y</code> or <code>Z</code> which then is emulated by additional
instructions.
For example, loading a value with <code>X+const</code> addressing with a
small non-negative <code>const &lt; 64</code> to a register <var>Rn</var> is
performed as
<pre class="example"> adiw r26, const ; X += const
ld <var>Rn</var>, X ; <var>Rn</var> = *X
sbiw r26, const ; X -= const
</pre>
<br><dt><code>-mtiny-stack</code><dd><a name="index-mtiny_002dstack-1446"></a>Only change the lower 8&nbsp;bits of the stack pointer.
<br><dt><code>-mfract-convert-truncate</code><dd><a name="index-mfract_002dconvert_002dtruncate-1447"></a>Allow to use truncation instead of rounding towards zero for fractional fixed-point types.
<br><dt><code>-nodevicelib</code><dd><a name="index-nodevicelib-1448"></a>Don't link against AVR-LibC's device specific library <code>lib&lt;mcu&gt;.a</code>.
<br><dt><code>-Waddr-space-convert</code><dd><a name="index-Waddr_002dspace_002dconvert-1449"></a>Warn about conversions between address spaces in the case where the
resulting address space is not contained in the incoming address space.
<br><dt><code>-Wmisspelled-isr</code><dd><a name="index-Wmisspelled_002disr-1450"></a>Warn if the ISR is misspelled, i.e. without __vector prefix.
Enabled by default.
</dl>
<h5 class="subsubsection">3.17.5.1 <code>EIND</code> and Devices with More Than 128 Ki Bytes of Flash</h5>
<p><a name="index-g_t_0040code_007bEIND_007d-1451"></a>Pointers in the implementation are 16&nbsp;bits wide.
The address of a function or label is represented as word address so
that indirect jumps and calls can target any code address in the
range of 64&nbsp;Ki words.
<p>In order to facilitate indirect jump on devices with more than 128&nbsp;Ki
bytes of program memory space, there is a special function register called
<code>EIND</code> that serves as most significant part of the target address
when <code>EICALL</code> or <code>EIJMP</code> instructions are used.
<p>Indirect jumps and calls on these devices are handled as follows by
the compiler and are subject to some limitations:
<ul>
<li>The compiler never sets <code>EIND</code>.
<li>The compiler uses <code>EIND</code> implicitely in <code>EICALL</code>/<code>EIJMP</code>
instructions or might read <code>EIND</code> directly in order to emulate an
indirect call/jump by means of a <code>RET</code> instruction.
<li>The compiler assumes that <code>EIND</code> never changes during the startup
code or during the application. In particular, <code>EIND</code> is not
saved/restored in function or interrupt service routine
prologue/epilogue.
<li>For indirect calls to functions and computed goto, the linker
generates <em>stubs</em>. Stubs are jump pads sometimes also called
<em>trampolines</em>. Thus, the indirect call/jump jumps to such a stub.
The stub contains a direct jump to the desired address.
<li>Linker relaxation must be turned on so that the linker generates
the stubs correctly in all situations. See the compiler option
<samp><span class="option">-mrelax</span></samp> and the linker option <samp><span class="option">--relax</span></samp>.
There are corner cases where the linker is supposed to generate stubs
but aborts without relaxation and without a helpful error message.
<li>The default linker script is arranged for code with <code>EIND = 0</code>.
If code is supposed to work for a setup with <code>EIND != 0</code>, a custom
linker script has to be used in order to place the sections whose
name start with <code>.trampolines</code> into the segment where <code>EIND</code>
points to.
<li>The startup code from libgcc never sets <code>EIND</code>.
Notice that startup code is a blend of code from libgcc and AVR-LibC.
For the impact of AVR-LibC on <code>EIND</code>, see the
<a href="http://nongnu.org/avr-libc/user-manual/">AVR-LibC&nbsp;user&nbsp;manual</a><!-- /@w -->.
<li>It is legitimate for user-specific startup code to set up <code>EIND</code>
early, for example by means of initialization code located in
section <code>.init3</code>. Such code runs prior to general startup code
that initializes RAM and calls constructors, but after the bit
of startup code from AVR-LibC that sets <code>EIND</code> to the segment
where the vector table is located.
<pre class="example"> #include &lt;avr/io.h&gt;
static void
__attribute__((section(".init3"),naked,used,no_instrument_function))
init3_set_eind (void)
{
__asm volatile ("ldi r24,pm_hh8(__trampolines_start)\n\t"
"out %i0,r24" :: "n" (&amp;EIND) : "r24","memory");
}
</pre>
<p class="noindent">The <code>__trampolines_start</code> symbol is defined in the linker script.
<li>Stubs are generated automatically by the linker if
the following two conditions are met:
<ul>
<li>The address of a label is taken by means of the <code>gs</code> modifier
(short for <em>generate stubs</em>) like so:
<pre class="example"> LDI r24, lo8(gs(<var>func</var>))
LDI r25, hi8(gs(<var>func</var>))
</pre>
<li>The final location of that label is in a code segment
<em>outside</em> the segment where the stubs are located.
</ul>
<li>The compiler emits such <code>gs</code> modifiers for code labels in the
following situations:
<ul>
<li>Taking address of a function or code label.
<li>Computed goto.
<li>If prologue-save function is used, see <samp><span class="option">-mcall-prologues</span></samp>
command-line option.
<li>Switch/case dispatch tables. If you do not want such dispatch
tables you can specify the <samp><span class="option">-fno-jump-tables</span></samp> command-line option.
<li>C and C++ constructors/destructors called during startup/shutdown.
<li>If the tools hit a <code>gs()</code> modifier explained above.
</ul>
<li>Jumping to non-symbolic addresses like so is <em>not</em> supported:
<pre class="example"> int main (void)
{
/* Call function at word address 0x2 */
return ((int(*)(void)) 0x2)();
}
</pre>
<p>Instead, a stub has to be set up, i.e. the function has to be called
through a symbol (<code>func_4</code> in the example):
<pre class="example"> int main (void)
{
extern int func_4 (void);
/* Call function at byte address 0x4 */
return func_4();
}
</pre>
<p>and the application be linked with <samp><span class="option">-Wl,--defsym,func_4=0x4</span></samp>.
Alternatively, <code>func_4</code> can be defined in the linker script.
</ul>
<h5 class="subsubsection">3.17.5.2 Handling of the <code>RAMPD</code>, <code>RAMPX</code>, <code>RAMPY</code> and <code>RAMPZ</code> Special Function Registers</h5>
<p><a name="index-g_t_0040code_007bRAMPD_007d-1452"></a><a name="index-g_t_0040code_007bRAMPX_007d-1453"></a><a name="index-g_t_0040code_007bRAMPY_007d-1454"></a><a name="index-g_t_0040code_007bRAMPZ_007d-1455"></a>Some AVR devices support memories larger than the 64&nbsp;KiB range
that can be accessed with 16-bit pointers. To access memory locations
outside this 64&nbsp;KiB range, the contentent of a <code>RAMP</code>
register is used as high part of the address:
The <code>X</code>, <code>Y</code>, <code>Z</code> address register is concatenated
with the <code>RAMPX</code>, <code>RAMPY</code>, <code>RAMPZ</code> special function
register, respectively, to get a wide address. Similarly,
<code>RAMPD</code> is used together with direct addressing.
<ul>
<li>The startup code initializes the <code>RAMP</code> special function
registers with zero.
<li>If a <a href="AVR-Named-Address-Spaces.html#AVR-Named-Address-Spaces">named address space</a> other than
generic or <code>__flash</code> is used, then <code>RAMPZ</code> is set
as needed before the operation.
<li>If the device supports RAM larger than 64&nbsp;KiB and the compiler
needs to change <code>RAMPZ</code> to accomplish an operation, <code>RAMPZ</code>
is reset to zero after the operation.
<li>If the device comes with a specific <code>RAMP</code> register, the ISR
prologue/epilogue saves/restores that SFR and initializes it with
zero in case the ISR code might (implicitly) use it.
<li>RAM larger than 64&nbsp;KiB is not supported by GCC for AVR targets.
If you use inline assembler to read from locations outside the
16-bit address range and change one of the <code>RAMP</code> registers,
you must reset it to zero after the access.
</ul>
<h5 class="subsubsection">3.17.5.3 AVR Built-in Macros</h5>
<p>GCC defines several built-in macros so that the user code can test
for the presence or absence of features. Almost any of the following
built-in macros are deduced from device capabilities and thus
triggered by the <samp><span class="option">-mmcu=</span></samp> command-line option.
<p>For even more AVR-specific built-in macros see
<a href="AVR-Named-Address-Spaces.html#AVR-Named-Address-Spaces">AVR Named Address Spaces</a> and <a href="AVR-Built_002din-Functions.html#AVR-Built_002din-Functions">AVR Built-in Functions</a>.
<dl>
<dt><code>__AVR_ARCH__</code><dd>Build-in macro that resolves to a decimal number that identifies the
architecture and depends on the <samp><span class="option">-mmcu=</span><var>mcu</var></samp> option.
Possible values are:
<p><code>2</code>, <code>25</code>, <code>3</code>, <code>31</code>, <code>35</code>,
<code>4</code>, <code>5</code>, <code>51</code>, <code>6</code>
<p>for <var>mcu</var>=<code>avr2</code>, <code>avr25</code>, <code>avr3</code>, <code>avr31</code>,
<code>avr35</code>, <code>avr4</code>, <code>avr5</code>, <code>avr51</code>, <code>avr6</code>,
<p>respectively and
<p><code>100</code>,
<code>102</code>, <code>103</code>, <code>104</code>,
<code>105</code>, <code>106</code>, <code>107</code>
<p>for <var>mcu</var>=<code>avrtiny</code>,
<code>avrxmega2</code>, <code>avrxmega3</code>, <code>avrxmega4</code>,
<code>avrxmega5</code>, <code>avrxmega6</code>, <code>avrxmega7</code>, respectively.
If <var>mcu</var> specifies a device, this built-in macro is set
accordingly. For example, with <samp><span class="option">-mmcu=atmega8</span></samp> the macro is
defined to <code>4</code>.
<br><dt><code>__AVR_</code><var>Device</var><code>__</code><dd>Setting <samp><span class="option">-mmcu=</span><var>device</var></samp> defines this built-in macro which reflects
the device's name. For example, <samp><span class="option">-mmcu=atmega8</span></samp> defines the
built-in macro <code>__AVR_ATmega8__</code>, <samp><span class="option">-mmcu=attiny261a</span></samp> defines
<code>__AVR_ATtiny261A__</code>, etc.
<p>The built-in macros' names follow
the scheme <code>__AVR_</code><var>Device</var><code>__</code> where <var>Device</var> is
the device name as from the AVR user manual. The difference between
<var>Device</var> in the built-in macro and <var>device</var> in
<samp><span class="option">-mmcu=</span><var>device</var></samp> is that the latter is always lowercase.
<p>If <var>device</var> is not a device but only a core architecture like
&lsquo;<samp><span class="samp">avr51</span></samp>&rsquo;, this macro is not defined.
<br><dt><code>__AVR_DEVICE_NAME__</code><dd>Setting <samp><span class="option">-mmcu=</span><var>device</var></samp> defines this built-in macro to
the device's name. For example, with <samp><span class="option">-mmcu=atmega8</span></samp> the macro
is defined to <code>atmega8</code>.
<p>If <var>device</var> is not a device but only a core architecture like
&lsquo;<samp><span class="samp">avr51</span></samp>&rsquo;, this macro is not defined.
<br><dt><code>__AVR_XMEGA__</code><dd>The device / architecture belongs to the XMEGA family of devices.
<br><dt><code>__AVR_HAVE_ELPM__</code><dd>The device has the the <code>ELPM</code> instruction.
<br><dt><code>__AVR_HAVE_ELPMX__</code><dd>The device has the <code>ELPM R</code><var>n</var><code>,Z</code> and <code>ELPM
R</code><var>n</var><code>,Z+</code> instructions.
<br><dt><code>__AVR_HAVE_MOVW__</code><dd>The device has the <code>MOVW</code> instruction to perform 16-bit
register-register moves.
<br><dt><code>__AVR_HAVE_LPMX__</code><dd>The device has the <code>LPM R</code><var>n</var><code>,Z</code> and
<code>LPM R</code><var>n</var><code>,Z+</code> instructions.
<br><dt><code>__AVR_HAVE_MUL__</code><dd>The device has a hardware multiplier.
<br><dt><code>__AVR_HAVE_JMP_CALL__</code><dd>The device has the <code>JMP</code> and <code>CALL</code> instructions.
This is the case for devices with more than 8&nbsp;KiB of program
memory.
<br><dt><code>__AVR_HAVE_EIJMP_EICALL__</code><dt><code>__AVR_3_BYTE_PC__</code><dd>The device has the <code>EIJMP</code> and <code>EICALL</code> instructions.
This is the case for devices with more than 128&nbsp;KiB of program memory.
This also means that the program counter
(PC) is 3&nbsp;bytes wide.
<br><dt><code>__AVR_2_BYTE_PC__</code><dd>The program counter (PC) is 2&nbsp;bytes wide. This is the case for devices
with up to 128&nbsp;KiB of program memory.
<br><dt><code>__AVR_HAVE_8BIT_SP__</code><dt><code>__AVR_HAVE_16BIT_SP__</code><dd>The stack pointer (SP) register is treated as 8-bit respectively
16-bit register by the compiler.
The definition of these macros is affected by <samp><span class="option">-mtiny-stack</span></samp>.
<br><dt><code>__AVR_HAVE_SPH__</code><dt><code>__AVR_SP8__</code><dd>The device has the SPH (high part of stack pointer) special function
register or has an 8-bit stack pointer, respectively.
The definition of these macros is affected by <samp><span class="option">-mmcu=</span></samp> and
in the cases of <samp><span class="option">-mmcu=avr2</span></samp> and <samp><span class="option">-mmcu=avr25</span></samp> also
by <samp><span class="option">-msp8</span></samp>.
<br><dt><code>__AVR_HAVE_RAMPD__</code><dt><code>__AVR_HAVE_RAMPX__</code><dt><code>__AVR_HAVE_RAMPY__</code><dt><code>__AVR_HAVE_RAMPZ__</code><dd>The device has the <code>RAMPD</code>, <code>RAMPX</code>, <code>RAMPY</code>,
<code>RAMPZ</code> special function register, respectively.
<br><dt><code>__NO_INTERRUPTS__</code><dd>This macro reflects the <samp><span class="option">-mno-interrupts</span></samp> command-line option.
<br><dt><code>__AVR_ERRATA_SKIP__</code><dt><code>__AVR_ERRATA_SKIP_JMP_CALL__</code><dd>Some AVR devices (AT90S8515, ATmega103) must not skip 32-bit
instructions because of a hardware erratum. Skip instructions are
<code>SBRS</code>, <code>SBRC</code>, <code>SBIS</code>, <code>SBIC</code> and <code>CPSE</code>.
The second macro is only defined if <code>__AVR_HAVE_JMP_CALL__</code> is also
set.
<br><dt><code>__AVR_ISA_RMW__</code><dd>The device has Read-Modify-Write instructions (XCH, LAC, LAS and LAT).
<br><dt><code>__AVR_SFR_OFFSET__=</code><var>offset</var><dd>Instructions that can address I/O special function registers directly
like <code>IN</code>, <code>OUT</code>, <code>SBI</code>, etc. may use a different
address as if addressed by an instruction to access RAM like <code>LD</code>
or <code>STS</code>. This offset depends on the device architecture and has
to be subtracted from the RAM address in order to get the
respective I/O&nbsp;address.
<br><dt><code>__AVR_SHORT_CALLS__</code><dd>The <samp><span class="option">-mshort-calls</span></samp> command line option is set.
<br><dt><code>__AVR_PM_BASE_ADDRESS__=</code><var>addr</var><dd>Some devices support reading from flash memory by means of <code>LD*</code>
instructions. The flash memory is seen in the data address space
at an offset of <code>__AVR_PM_BASE_ADDRESS__</code>. If this macro
is not defined, this feature is not available. If defined,
the address space is linear and there is no need to put
<code>.rodata</code> into RAM. This is handled by the default linker
description file, and is currently available for
<code>avrtiny</code> and <code>avrxmega3</code>. Even more convenient,
there is no need to use address spaces like <code>__flash</code> or
features like attribute <code>progmem</code> and <code>pgm_read_*</code>.
<br><dt><code>__WITH_AVRLIBC__</code><dd>The compiler is configured to be used together with AVR-Libc.
See the <samp><span class="option">--with-avrlibc</span></samp> configure option.
</dl>
</body></html>