Add a tool to modify control features in ELF binaries.
Submitted by: Bora Ozarslan borako.ozarslan@gmail.com
Details
Details
- Reviewers
borako.ozarslan_gmail.com - Group Reviewers
manpages - Commits
- rS352797: Add tool to modify ELF binary feature control bits
Diff Detail
Diff Detail
- Repository
- rS FreeBSD src repository - subversion
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
| tools/tools/controlelf/Makefile | ||
|---|---|---|
| 4 ↗ | (On Diff #54208) | This is not needed, right ? |
| 6 ↗ | (On Diff #54208) | And this. |
| 23 ↗ | (On Diff #54208) | Why this is needed ? Perhaps you should use #include "" instead of <>. |
| 24 ↗ | (On Diff #54208) | And what is this ? |
| 27 ↗ | (On Diff #54208) | Bump WARNS to max, and perhaps make warnings an error. New code should be clean. |
| tools/tools/controlelf/controlelf.1 | ||
| 1 ↗ | (On Diff #54208) | Really ? |
| tools/tools/controlelf/controlelf.c | ||
| 160 ↗ | (On Diff #54208) | Extra blank line. |
| 161 ↗ | (On Diff #54208) | if (elf != NULL) |
| 164 ↗ | (On Diff #54208) | Checking for error from close(2) is almost always pointless, perhaps except for EBADF. |
| 186 ↗ | (On Diff #54208) | Blank line after '{' if no locals. |
| 191 ↗ | (On Diff #54208) | Use bool's ? |
| 199 ↗ | (On Diff #54208) | return (0); |
| 200 ↗ | (On Diff #54208) | I do not understand this loop, in particular the condition inside the if(). |
| 212 ↗ | (On Diff #54208) | No initializers in local declarations. |
| 226 ↗ | (On Diff #54208) | nitems() |
| 245 ↗ | (On Diff #54208) | Excess (). |
| 261 ↗ | (On Diff #54208) | return on the next line. |
| 315 ↗ | (On Diff #54208) | Check for error from lseek. |
| 319 ↗ | (On Diff #54208) | Check errors from read, check for short reads. |
| 326 ↗ | (On Diff #54208) | roundup2(). |
| 327 ↗ | (On Diff #54208) | sizeof(char) is sillyness. Why do you need calloc at all ? |
| 336 ↗ | (On Diff #54208) | This is yet another incorrect use of strncmp. At least you must check that strlen(name) == 7. |
| 345 ↗ | (On Diff #54208) | Why this cast ? It looks arbitrary. |
| 362 ↗ | (On Diff #54208) | branch statement on the next line. |
| tools/tools/controlelf/controlelf.1 | ||
|---|---|---|
| 24–25 ↗ | (On Diff #54208) | These two lines should be replaced with just: .\" $FreeBSD$ |
| tools/tools/controlelf/controlelf.c | ||
|---|---|---|
| 200 ↗ | (On Diff #54208) | Essentially the for loop is going through each program header and finding the one that is of type PT_NOTE and returning that. In libelf, the recommended way to enumerate through the program headers is this way. The program header is returned inside the phdr which is given as a parameter. The function gelf_getphdr gets the ith program header and replaces the values inside the phdr. On success it returns phdr (the address given to it) and on failure it returns NULL. So that's what the first if is checking for, if the function succeeded or not. Then the second if checks if this program header is of type PT_NOTE. I now see that there may be more than 1 PT_NOTE header. |
Comment Actions
Fixed Makefile.
Changed code to comply with style(9).
Some cleanup.
Check all PT_NOTE headers until one that has feature control note has been found.
Updated copyright notices.
Get rid of control features that haven't been implemented yet (they are in review). Get rid of all mentions/examples of it.