Hi,
> > +int globErrFunc(const char *epath, int eerrno)
> > +{
> > + /* TODO check fatal errors */
> > +
> > + return 0;
> > +}
>
> I think pjones's comments from
> https://www.redhat.com/archives/anaconda-devel-list/2009-December/msg00421.html
> still apply here.
No they don't as I said earlier. I _want_ to ignore most of the errors. The only errors we need to check here are out of memory and similar ones. And that is why I marked this TODO before.
return LOADER_BACK;
> > + for (fnPtr = driverDiskFiles; *fnPtr; fnPtr++) {
> > + snprintf(file, 200, "%s/%s/%s", mntpt, getProductArch(),
> *fnPtr);
> > + if (access(file, R_OK)) {
> > + logMessage(ERROR, "cannot find %s, bad driver disk",
> file);
> > + return LOADER_BACK;
> > + }
> > + }
> >
> > return LOADER_OK;
> > }
>
> Likewise here.
Well pjones' comment said:
"I realize mntpt and getProductArch's return are unlikely to be very large, but
these should still probably be snprintf() or checked_asprintf() (... or sprintfa()
if it were to suddenly come into existence ;)"
And I see snprintf in the code. The checking could be more thorough, that is true. But it does use space limited version of sprintf.
Martin
_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
