| View previous topic :: View next topic |
| Author |
Message |
bryant@balancebitsconsult
Joined: 21 Nov 2023
Posts: 22
|
| COPY_ARRAY Macro |
 Posted: Thu Dec 07, 2023 6:45 pm |
 |
|
So, there seems to be a COPY_ARRAY() macro. Undocumented ?
Anyone know what it does ? How it works ? I am running into it. |
|
 |
Ttelmah
Joined: 11 Mar 2010
Posts: 19244
|
|
| Posted: Fri Dec 08, 2023 3:03 am |
|
|
I don't think it does.
I think you are including something that defines this, and are then trying
to define a second copy.
Possibly you have included a library from some other source that defines
this?.
I Just tried creating a macro with this name on a few different projects,
and none complain that I am redefining a name, and the macros all work
OK.
What is happening?.
What error are you getting?.
I seem to remember MicroChip having a copy_array function or macro in
one of their libraries. |
|
|
bryant@balancebitsconsult
Joined: 21 Nov 2023
Posts: 22
|
|
| Posted: Fri Dec 08, 2023 10:54 am |
|
|
| Ttelmah wrote: | I don't think it does.
I think you are including something that defines this, and are then trying
to define a second copy.
Possibly you have included a library from some other source that defines
this?.
I Just tried creating a macro with this name on a few different projects,
and none complain that I am redefining a name, and the macros all work
OK.
What is happening?.
What error are you getting?.
I seem to remember MicroChip having a copy_array function or macro in
one of their libraries. |
I have added it to a repo to examine. I was seeing errors when disabling case insensitivity at the top of util.h (link) : https://github.com/beadon/circbuf_demo/blob/main/util.h
Now, strangely, I am unable to reproduce the errors triggered by the case change! So, I will take further steps to reproduce and reply on this thread. I have a separate issue in defining the macro - it MIGHT be related --
errors are :
| Code: |
Compiling C:\Users\beadon\Documents\GitHub\circbuf_demo\main on 08-Dec-23 at 08:49
*** Error 9 "C:\Users\beadon\Documents\GitHub\circbuf_demo\util.h" Line 95(31,37): Macro identifier requires parameters
*** Error 12 "C:\Users\beadon\Documents\GitHub\circbuf_demo\util.h" Line 95(37,38): Undefined identifier val
2 Errors, 0 Warnings.
Build Failed.
|
Which, at a glance, seems trivial to resolve. Once you look at the line though, there's no "val" and the error message becomes more puzzling.
| Code: | | #define unsigned_mult_overflows(a, b) ((a) && (b) > maximum_unsigned_value_of_type(a) / (a)) |
Any insights are appreciated. |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19244
|
|
| Posted: Fri Dec 08, 2023 11:29 am |
|
|
Two things.
The first is what is line 95 (and a couple of lines in front)?.
The second ties to the second part of the above. The compilers fault
identifier code has a habit of misidentifying problems. It goes wrong several
lines after the actual fault, when it can no longer carry on. So an
unidentified fault gets reported as a completely different fault lots later.
This is why sometimes using a C syntax scanner tool can be a way of
finding the real problem. |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19244
|
|
| Posted: Fri Dec 08, 2023 12:43 pm |
|
|
Change the word 'unsigned' in your value there to something like 'uns'.
unsigned is a keyword, and in general keywords should not be used like this.
Looking at it more. I wonder how maximum_unsigned_value_of_type is
coded?. My guess would be that it is a macro, that possibly uses sizeof
to read a value from an array containing the size limits for the integer
types. You have a significant danger when coding macros inside macros,
that you can take the resulting line length over the maximum supported
by the compiler.
What overflow are you actually trying to test?.
| Code: |
//check whether a*x could overflow
// If one number is -1 and another is INT_MIN, multiplying them we get
// abs(INT_MIN) which is 1 higher than INT_MAX
if (a == -1 && x == INT_MIN) // `a * x` can overflow
if (x == -1 && a == INT_MIN) // `a * x` (or `a / x`) can overflow
// general case
if (x != 0 && a > INT_MAX / x) // `a * x` would overflow
if (x != 0 && a < INT_MIN / x) // `a * x` would underflow
|
Using your macro for the MAX, you should bracket round this since
otherwise the expansion may not be as you expect. |
|
|
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
