Databases
StackOverflow2013
Postsdbo
POFlow controlling macros with 'goto'
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POFlow controlling macros with 'goto'
    primarykey
    data
    text
    <p>Yes, <strong>two hated constructs combined</strong>. Is it as bad as it sounds or can it be seen as a good way to control usage of goto and also provide a reasonable cleanup strategy?</p> <p>At work we had a discussion about whether or not to allow goto in our coding standard. In general nobody wanted to allow free usage of goto but some were positive about using it for cleanup jumps. As in this code:</p> <pre><code>void func() { char* p1 = malloc(16); if( !p1 ) goto cleanup; char* p2 = malloc(16); if( !p2 ) goto cleanup; goto norm_cleanup; err_cleanup: if( p1 ) free(p1); if( p2 ) free(p2); norm_cleanup: } </code></pre> <p>The abovious benefit of such use is that you don't have to end up with this code:</p> <pre><code>void func() { char* p1 = malloc(16); if( !p1 ){ return; } char* p2 = malloc(16); if( !p2 ){ free(p1); return; } char* p3 = malloc(16); if( !p3 ){ free(p1); free(p2); return; } } </code></pre> <p>Especially in constructor-like functions with many allocations this can sometimes grow very bad, not the least when someone has to insert something in the middle.</p> <p>So, in order to be able to use goto, but still clearly isolate it from being used freely, a set of flow controlling macros was created for handling the task. Looks something like this (simplified):</p> <pre><code>#define FAIL_SECTION_BEGIN int exit_code[GUID] = 0; #define FAIL_SECTION_DO_EXIT_IF( cond, exitcode ) if(cond){exit_code[GUID] = exitcode; goto exit_label[GUID];} #define FAIL_SECTION_ERROR_EXIT(code) exit_label[GUID]: if(exit_code[GUID]) int code = exit_code[GUID];else goto end_label[GUID] #define FAIL_SECTION_END end_label[GUID]: </code></pre> <p>We can use this as follows:</p> <pre><code>int func() { char* p1 = NULL; char* p2 = NULL; char* p3 = NULL; FAIL_SECTION_BEGIN { p1 = malloc(16); FAIL_SECTION_DO_EXIT_IF( !p1, -1 ); p2 = malloc(16); FAIL_SECTION_DO_EXIT_IF( !p2, -1 ); p3 = malloc(16); FAIL_SECTION_DO_EXIT_IF( !p3, -1 ); } FAIL_SECTION_ERROR_EXIT( code ) { if( p3 ) free(p3); if( p2 ) free(p2); if( p1 ) free(p1); return code; } FAIL_SECTION_END return 0; </code></pre> <p>It looks nice, and comes with many benefits, BUT, are there any drawbacks we should be thinking about before rolling this out into development? It is after all very flow controlling and goto:ish. Both are discouraged. What are the arguments for discouraging them in this case?</p> <p>Thanks.</p>
    singulars
    1. PO
      singulars
      1. PTAnswer
    1. This table or related slice is empty.
    1. PTQuestion
    1. USsharkin
    1. USsharkin
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    2. PO
      singulars
      1. PTAnswer
    3. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POFlow controlling macros with 'goto'
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POFlow controlling macros with 'goto'
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. POFlow controlling macros with 'goto'
      1. This table or related slice is empty.
      1. VTUpMod
    1. COAs Martin Fido points out in his answer, in the final code snippet the pointers p1, p2 and p3 are all out of inside the error-handling section, and in the first code snippet p2 will contain rubbish data (if it compiles at all -- not sure about the C rules).
      singulars
      1. POFlow controlling macros with 'goto'
      1. USj_random_hacker
    2. CO@j: I'm sure this code won't even go through the compiler. Thanks for the tip, but you must have missed the entire point of the question.
      singulars
      1. POFlow controlling macros with 'goto'
      1. USsharkin
    3. CO@R.A: What makes you think I missed the point? The non-compilability was a confusing side-issue that I thought was worth pointing out in a comment.
      singulars
      1. POFlow controlling macros with 'goto'
      1. USj_random_hacker
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload