The following example shows cases that were mis-evaluated:
/* Should print "3 10000" */
#include <stdio.h>
int main(void)
{
printf("%lu %lu\n", 100000ul / 30000ul, 100000ul % 30000ul);
}
This is a problem introduced by the scanner changes between ORCA/C 2.1.0 and ORCA/C 2.1.1 B3.
The following examples demonstrate the problem:
#define m m
m
#define f(x) f(x)
f(a)
This occurred because the values were being rounded rather than truncated when converted to long, unsigned long, or unsigned int.
This was causing problems in the C6.2.3.5.CC test case when compiled with optimization.
The below program demonstrates the problem:
#pragma optimize 1
#include <stdio.h>
int main (void)
{
long L;
unsigned int ui;
unsigned long ul;
L = -1.5;
ui = 1.5;
ul = 1.5;
printf("%li %u %lu\n", L, ui, ul); /* should print "-1 1 1" */
}
This fixes the compco09.c test case.
This implementation permits duplicate copies of type qualifiers to appear. This is technically illegal in C90, but it’s legal in C99 and later, and ORCA/C already allows this in other contexts.
This allows functions that require an OMF segment byte count of up to 128K to be compiled, although the length in memory at run time is still limited to 64K. (The OMF segment byte count is usually larger, due to the size of relocation records, etc.)
This is useful for compiling large functions, e.g. the main interpreter loop in git. It also fixes the bug shown in the compca23 test case, where functions that require a segment of over 64K may appear to compile correctly but generate corrupted OMF segment headers. This related to tracking sizes with 16-bit values that could roll over.
This patch increases the memory needed at run time by 64K. This shouldn’t generally be a problem on systems with sufficient memory, although it does increase the minimum memory requirement a bit. If behavior in low-memory configurations is a concern, buffSize could be made into a run-time option.
This would occur if ORCA/C remained in memory and was restarted after a previous execution, because the 'pc' value was not reinitialized. The ORCA linker seems to ignore the too-long segment length value, but ORCA/C should generate a correct value that actually corresponds to the length of the segment.
This is necessary to compile some very large functions, such as the main interpreter loop in Git.
This consumes about 8K of extra memory for the additional label records.
The issue was that 16-bit absolute addressing (in the data bank) was being used to access the data to compare, but with the large memory model the static arrays or structs are not necessarily in the same bank, so absolute long addressing should be used.
This was sometimes causing failures in the C4.6.4.1.CC and C4.6.6.1.CC conformance tests in the ORCA/C test suite.
The following program often demonstrates the problem (depending on memory layout and contents):
#pragma memorymodel 1
#pragma optimize 1
#include <stdio.h>
int i;
char ch1[32000];
long L1[1];
int main (void)
{
if (L1 [0] != 0)
printf("%li\n", L1[0]); /* shouldn't print */
/* buggy behavior can happen if the bank bytes of these pointers differ */
printf("%p %p\n", &L1[0], &i);
}
This could cause problems when asm blocks contained instructions that the ORCA/C native code optimizer didn’t know about, as in the example below. It might also be possible to trigger this bug without asm blocks (particularly with the large memory model), but I haven’t run into a case that does.
The new approach conservatively assumes that unknown instructions block the optimization. This should be equivalent to the old code with respect to the instructions defined in CGI.pas, except that m_bit_imm should have been treated as blocking the optimization but was not. There are still some other potential problem cases with applying this lda-elimination optimization to arbitrary assembly code, but fixing them might interfere with the optimization in useful cases, so I’m leaving those alone for now.
Here is an example of a program with an asm block affected by this problem:
#pragma optimize 74
int x,y;
/* should print 2 when invoked with argc==1 */
int main(int argc, char **argv)
{
x = argc;
y = argc + 6;
asm {
lda #1
pha
eor >x
bne done
inc argc
done: pla
}
printf("%i\n", argc);
}
The issue was that one of the procedures used for CSE would recursively call itself for every 'next' link in the code of the basic block. To avoid this, I made it loop back to the top instead (i.e. did a manual tail-call elimination transformation).
This problem could be observed with large switch statements as in the following example, although other codes with very large basic blocks might have triggered it too. Whether ORCA/C actually crashes will depend on the memory layout--in my testing, this example consistently caused it to crash when running under GNO:
#pragma optimize 16
int main (int argc, char **argv)
{
switch (argc)
{
case 0: case 1: case 2: case 3: case 4: case 5: case 6: case 7:
case 8: case 9: case 10: case 11: case 12: case 13: case 14: case 15:
case 16: case 17: case 18: case 19: case 20: case 21: case 22: case 23:
case 24: case 25: case 26: case 27: case 28: case 29: case 30: case 31:
case 32: case 33: case 34: case 35: case 36: case 37: case 38: case 39:
case 40: case 41: case 42:
case 262:
;
}
}
Also, fix a case where an uninitialized value could be used, potentially resulting in errors not being reported (although I haven’t seen that in practice).
This fixes problems where >>= operations might not use an arithmetic shift in certain cases where they should, as in the below program:
#include <stdio.h>
int main (void)
{
int i;
unsigned u;
long l;
unsigned long ul;
i = -1;
u = 1;
i >>= u;
printf("%i\n", i); /* should be -1 */
l = -1;
ul = 3;
l >>= ul;
printf("%li\n", l); /* should be -1 */
}
This is necessary so that subsequent processing sees the correct expression type and proceeds accordingly. In particular, without this patch, the casts in the below program were erroneously ignored:
#include <stdio.h>
int main(void)
{
unsigned int u;
unsigned char c;
c = 0;
u = (unsigned char)~c;
printf("%u\n", u);
c = 200;
printf("%i\n", (unsigned char)(c+c));
}
This is as required by the C standards: the type of the right operand should not affect the result type.
The following program demonstrates problems with the old behavior:
#include <stdio.h>
int main(void)
{
unsigned long ul;
long l;
unsigned u;
int i;
ul = 0x8000 << 1L; /* should be 0 */
printf("%lx\n", ul);
l = -1 >> 1U; /* should be -1 */
printf("%ld\n", l);
u = 0xFF10;
l = 8;
ul = u << l; /* should be 0x1000 */
printf("%lx\n", ul);
l = -4;
ul = 1;
l = l >> ul; /* should be -2 */
printf("%ld\n", l);
}
The following demonstrates cases that would erroneously be allowed (and misleadingly give a size of 0) before:
#include <stdio.h>
struct s *S;
int main(void)
{
printf("%lu %lu\n", sizeof(struct s), sizeof *S);
}
The following test case demonstrates the problem:
#include <stdio.h>
int main (void)
{
if (sizeof(int) - 5 < 0) puts("error 1");
if (sizeof &main - 9 < 0) puts("error 2");
}
This affected binary or unary expressions with constant operands, at least one of which was of type unsigned long.
The following test case demonstrates the problem:
#include <stdio.h>
int main (void)
{
if (0 + 0x80000000ul < 0) puts("error 1");
if (~0ul < 0) puts("error 2");
}
This cuts a few instructions from code like what is shown in commit affbe9. (It also works around the bug with that example, although that patch addresses the root cause of the problem.)
This generated invalid code in instances like the following. The code generated for "s = s->u.next" would update the most significant word of s first, then use an indirect load with the half-updated pointer value to update the least significant word of s. This would generally corrupt the result if the new and old pointers had different bank bytes.
#pragma optimize 79
#include <stdio.h>
struct S {
int i;
union {
struct S * next;
} u;
} s1 = {0, 0};
int main (void)
{
struct S * s = &s1;
s = s->u.next;
if (s != 0)
puts("compiler bug detected\n"); /* May not always be triggered, depending on memory contents. */
}
ORCA/C previously allowed struct/union members to be declared with incomplete type. Because of this, it allowed C99-style flexible array members to be declared, albeit by accident rather than by design. In some basic testing, these seem to work correctly, except that they could be initialized and that would give rise to odd behavior.
I have restricted it to allowing flexible array members only in the cases allowed by C99/C11, and otherwise disallowing members with incomplete type. I have also prohibited initializing flexible array members.
Also, generate better code for zero-initializing small arrays.
The problem was that the code would call the library routine ~ZERO with a size of 1, but it only works properly with a size of 2 or more. While adding a check here, I also changed it to not call ~ZERO for other small arrays (<=10 bytes), since it is generally more efficient to just initialize them directly.
The initializations in the following are examples that could trigger the problem:
int main(void)
{
struct { int i; char s[1]; } foo = {1, 0};
char arr[2][1] = {2};
}
This may be someone trying to use a C11-style anonymous struct/union, which should be flagged as an error until and unless those are supported. Otherwise, it probably just indicates that the programmer is confused. In any case, an error should be flagged for it.
C89 restricts bit fields to (signed) int and unsigned int only, although later standards note that additional types may be supported. ORCA/C supports the other integer types as an extension.
This fixes the compco01.c test case.
These would previously be allowed, with the void value treated as if it was unsigned long.
Void values are still allowed for the second and third operands of the ?: operator, but only if they are both void. This is as required by the C standard.
This patch should also permit the union initialization code to handle unions containing bit fields, but for the time being they are still prohibited by code elsewhere in the compiler.
This fixes a problem with ORCA/C conformance test C5.6.0.1.CC, which was introduced by commit bf9fa66. A slightly more involved fix was needed to preserve the correct behavior while avoiding the memory trashing fixed by that patch.
Unsigned constants in the range 0x8000-0xFFFF were erroneously being treated like negative signed values in some contexts, including in array declarators and in the case labels of switch statements where the expression switched over has type long or unsigned long.
This could lead to bogus compile errors for array declarations and typedefs such as the following:
typedef char foo[0x8000];
It could also lead to cases in switch statements not being properly matched, as in the following program:
#include <stdio.h>
int main(void)
{
long i = 0xFF00;
switch (i) {
case 0xFF00:
puts("good");
break;
default:
puts("bad");
}
}
The case label values are converted to the promoted type of the expression being switched on, as if by a cast. In practice, this means discarding the high bits of a 32-bit value to produce a 16-bit one.
Code requiring this is dubious and would be a good candidate for a warning or a lint error, but it's allowed under the C standards.
The following code demonstrates the issue:
#include <stdio.h>
int main(void)
{
int i = 0x1234;
switch (i) {
case 0xABCD1234:
puts("good");
break;
default:
puts("bad");
}
}
This also fixes a logic error that may have permitted other conversions to be improperly omitted in some cases.
The following program demonstrates the problem (should print 211):
#pragma optimize 1
#include <stdio.h>
int main(void)
{
unsigned int i = 1234;
long l = (unsigned char)(i+1);
printf("%li\n", l);
}
This resulted from the addition of the signed-to-unsigned comparison optimization. Specifically, it calls TypeOf for the expressions on each side of the comparison, and this did not handle function calls. That support has now been added, and will give the proper return type for direct and indirect calls to C functions. The IR for tool calls doesn't include the return type (just the number of bytes), so we return cgVoid for them. This is OK for the present use case.
Without this fix, an expression of the form "0 * exp" would be reduced to simply "0" unless exp contained a function call; other side effects of exp (such as assignments or increments) would be removed.
A similar issue could occur with additions that use the same expression on both sides of the "+": after optimization, it would only be evaluated once. I think the cases addressed here are all undefined behavior under the C standards, so the old behavior wasn't technically wrong, but the new behavior is still less confusing.
Specifically, this ensures that the depth-first numbering of basic blocks starts from 1, which is what ReachingDefinitions expects. Without this fix, reaching definitions wouldn't be correctly computed for functions that contain unreachable basic blocks (including the implicit one to return at the end). This could result in invalid hoisting of operations out of the loop.
This fixes the compca26.c test case.
Specifically, convert signed word comparisons to unsigned if both sides were either unsigned byte values or non-negative constants. This is incorporated as part of intermediate code peephole optimization (bit 0).
This should alleviate some cases of performance regressions due to promoting char to int instead of unsigned int.
