Am 22. Januar 2021 09:54:20 MEZ schrieb Sughosh Ganu <sughosh.g...@linaro.org>: >On Thu, 21 Jan 2021 at 19:14, Heinrich Schuchardt <xypron.g...@gmx.de> >wrote: > >> On 21.01.21 12:36, Sughosh Ganu wrote: >> > >> > >> > On Thu, 21 Jan 2021 at 00:34, Tom Rini <tr...@konsulko.com >> > <mailto:tr...@konsulko.com>> wrote: >> > >> > I decided to run Coverity part-way through the merge window >this time >> > and here's what's been found so far. >> > >> > ----- Forwarded message from scan-ad...@coverity.com >> > <mailto:scan-ad...@coverity.com> ----- >> > >> > Date: Mon, 18 Jan 2021 17:53:19 +0000 (UTC) >> > From: scan-ad...@coverity.com <mailto:scan-ad...@coverity.com> >> > To: tom.r...@gmail.com <mailto:tom.r...@gmail.com> >> > Subject: New Defects reported by Coverity Scan for Das U-Boot >> > >> > Hi, >> > >> > Please find the latest report on new defect(s) introduced to >Das >> > U-Boot found with Coverity Scan. >> > >> > 23 new defect(s) introduced to Das U-Boot found with Coverity >Scan. >> > 2 defect(s), reported by Coverity Scan earlier, were marked >fixed in >> > the recent build analyzed by Coverity Scan. >> > >> > New defect(s) Reported-by: Coverity Scan >> > Showing 20 of 23 defect(s) >> > >> > ** CID 316356: Resource leaks (RESOURCE_LEAK) >> > /tools/mkeficapsule.c: 225 in add_public_key() >> > >> > >> > <snip> >> > >> > >> > >> > >> >________________________________________________________________________________________________________ >> > *** CID 316356: Resource leaks (RESOURCE_LEAK) >> > /tools/mkeficapsule.c: 225 in add_public_key() >> > 219 if (ret < 0) { >> > 220 fprintf(stderr, "%s: Unable to add >public >> > key to the FDT\n", >> > 221 __func__); >> > 222 goto err; >> > 223 } >> > 224 >> > >>> CID 316356: Resource leaks (RESOURCE_LEAK) >> > >>> Handle variable "srcfd" going out of scope leaks the >handle. >> > 225 return 0; >> > 226 >> > 227 err: >> > 228 if (sptr) >> > 229 munmap(sptr, src_size); >> > 230 >> > >> > >> > I think these should not cause any issues, since the function >return >> > results in the process termination in both the scenarios of success >and >> > failure. But i will post a patch to handle these errors to keep the >> > resource handling consistent. >> >> Looking at line 234f: >> >> if (srcfd >= 0) >> close(srcfd); >> >> The comparison is wrong. It should be: >> >> if (srcfd != -1) >> close(srcfd); >> >> The open.2 man-page says that only -1 signals an error. According to >the >> man-page -2 is a legal value for a file descriptor. >> > >Can you point me to which man page you are referring to. The open >manpage >on my ubuntu system has the following, > >"The return value of open() is a file descriptor, a small, nonnegative >integer that is used in subsequent system calls" > >I could not find any mention of -2 being a valid file descriptor. > >-sughosh
You are right https://pubs.opengroup.org/onlinepubs/9699919799/ says the return value must be positive or -1 (in case of an error).
