Youness Alaoui has posted comments on this change. ( https://review.coreboot.org/19066 )
Change subject: util/intelmetool: Fix access to deleted data on stack ......................................................................
Patch Set 1:
(3 comments)
https://review.coreboot.org/#/c/19066/1//COMMIT_MSG Commit Message:
PS1, Line 9: pci_me_interface_scan was returning (via argument 'name') a pointer : to the interface name which was stored in a stack variable. : This caused part of the name to be printed as garbage stack data : in some situations if stack data was overwritten.
There is no reason to break the line for a new sentence, if it’s in the sam
You mean between "stack variable." and "This caused" ? I broke that line because of the 80-char limit, not because it's a new sentence. I can merge 'this' into the previous line if that's what you want.
https://review.coreboot.org/#/c/19066/1/util/intelmetool/intelmetool.c File util/intelmetool/intelmetool.c:
PS1, Line 155: char
I think this should be `const char`. That libpci has a weird interface,
I don't think it should be a const char, since the value of name gets modified. Unless I'm misunderstanding what the const refers to in a 'const char **' declaration.
PS1, Line 238: name
It's never checked for a NULL-pointer, or did I miss that?
It's not checked, it's used for the printf below directly, but the value will always be set to either a static string or namebuf by pci_lookup_name, and pci_lookup_name is always called if 'dev' is returned non NULL, and 'dev' is checked. Either way, isn't printf NULL-safe when printing strings ?