Reduce number of warnings #179
Conversation
| @@ -45,7 +45,7 @@ class CTable | |||
|
|
|||
| CTable() | |||
| { | |||
| ::memset(_rgtype, 0, sizeof(_rgtype)); | |||
| ::memset((void*)_rgtype, 0, sizeof(_rgtype)); | |||
There was a problem hiding this comment.
Wait, this isn't fixing anything, it's just suppressing the warning... and the warning looks correct to me.
Utilities/socketxx/socket++/fork.cpp
Outdated
| { | ||
| ForkProcess* cur = Fork::ForkProcess::list; | ||
| while(cur != nullptr){ | ||
| ForkProcess* next = cur->next; | ||
| if(cur->kill_child) | ||
| delete cur; | ||
| break; | ||
| } | ||
|
|
||
| cur = next; | ||
| } | ||
There was a problem hiding this comment.
The commit message doesn't say what the change is about...
| @@ -607,7 +607,7 @@ class Element<TVR, VM::VM1_n> | |||
| Save = true; // ???? | |||
| if( Internal ) | |||
| { | |||
| memcpy(internal, Internal, len); | |||
| memcpy((void*)internal, (void*)Internal, Length); | |||
There was a problem hiding this comment.
Same comment here. This is suppressing a warning, not fixing anything...
Utilities/socketxx/socket++/fork.cpp
Outdated
| // First, kill all children whose kill_child flag is set. | ||
| // Second, wait for other children to die. | ||
| { | ||
| for (ForkProcess* cur = Fork::ForkProcess::list; cur; cur = cur->next) | ||
| if (cur->kill_child) | ||
|
|
||
| for (ForkProcess* cur = Fork::ForkProcess::list; cur != nullptr; cur = cur->next) | ||
| if (cur->kill_child) { | ||
| delete cur; | ||
| break; | ||
| } |
There was a problem hiding this comment.
The comment above this code says "kill all children" so breaking after killing one seems wrong...
| @@ -58,7 +58,7 @@ unique_ptr<JpegMarkerSegment> JpegMarkerSegment::CreateJpegFileInterchangeFormat | |||
| content.push_back(static_cast<uint8_t>(params.Ythumbnail)); | |||
| if (params.Xthumbnail > 0) | |||
| { | |||
| if (params.thumbnail) | |||
There was a problem hiding this comment.
This is 3rd party code, so needs separate handling.
Is this reversal of logic fixed upstream already perhaps?
There was a problem hiding this comment.
I am not sure if it's 3rd party code but may be a usage of 3rd party library.
Also, it does not seem to be a reversal of logic. This code sample was edited 8 years ago.
I made this change to align logic with the following exception message which is ""params.Xthumbnail is > 0 but params.thumbnail == null_ptr""
There was a problem hiding this comment.
I am not sure if it's 3rd party code but may be a usage of 3rd party library.
It definitely is: https://github.com/team-charls/charls
|
@MadVitaliy I've cherry-picked the two commits that were clear to me. Feel free to rebase and re-submitted with extended description. As explained third party update should not be addressed like this. You need to pull in the full version even in the case of a single line fix. Thanks ! |
Fix of: -Wclass-memaccess -Wuse-after-free -Werror=unused-but-set-variable -Werror=unused-variable -Werror=nonnull for GCC 12.
This PR is updated #177.