Added multiple NULL and nullptr checks#8962
Added multiple NULL and nullptr checks#8962Zakk0n wants to merge 1 commit intoFirebirdSQL:masterfrom
Conversation
aafemt
left a comment
aafemt
left a comment
There was a problem hiding this comment.
Mechanical "fixing" of AI-produced warnings, no code analyze performed. Too bad even for students.
| @@ -87,6 +87,8 @@ unsigned makeDynamicStrings(unsigned length, ISC_STATUS* const dst, const ISC_ST | |||
| case isc_arg_cstring: | |||
| fb_assert(string); | |||
| *to++ = (ISC_STATUS)(IPTR) string; | |||
| if (!string) | |||
There was a problem hiding this comment.
This is a dead code. The loop above makes sure that string is not nullptr.
| @@ -3367,7 +3368,8 @@ static void gen_t_start( const act* action, int column) | |||
|
|
|||
| if (trans->tra_db_count == 1) | |||
| { | |||
| printa(column, "%s = %s->startTransaction(%s, %d, fb_tpb_%d);", | |||
| if (trans->tra_tpb) | |||
There was a problem hiding this comment.
And what's else? Don't start the transaction at all?..
| @@ -990,15 +998,15 @@ void Applier::storeBlob(thread_db* tdbb, TraNumber traNum, bid* blobId, | |||
| fb_assert(blob->blb_flags & BLB_temporary); | |||
| fb_assert(!(blob->blb_flags & BLB_closed)); | |||
|
|
|||
| if (length) | |||
| if (blob && length) | |||
There was a problem hiding this comment.
...and if blob is nullptr - still get AV on BLB_close call. Instead else branch must be added to if in line 983 with bugcheck logging and bail.
| @@ -1069,6 +1049,9 @@ static Lock* hash_get_lock(Lock* lock, USHORT* hash_slot, Lock*** prior) | |||
|
|
|||
| // if no collisions found, we're done | |||
|
|
|||
| if (!att->att_compatibility_table) | |||
There was a problem hiding this comment.
att_compatibility_table is surely initialized in line 1042. Another dead code.
| @@ -1353,10 +1336,13 @@ static USHORT internal_downgrade(thread_db* tdbb, CheckStatusWrapper* statusVect | |||
|
|
|||
| // if we can convert to that level, set all identical locks as having that level | |||
|
|
|||
| if (!first || !first->lck_physical) | |||
| return NULL; | |||
There was a problem hiding this comment.
Not NULL but a proper LCK_* constant here. Which one is "proper"? That's the question. Investigation also can find out if cases when first is nullptr are possible at all. BTW, lck_physical is not a pointer.
|
How is this PR different from #8964 ? |
3 participants
This PR is a part of #8938