hardenedbsd/HardenedBSD
1
0
Fork 0
mirror of https://git.hardenedbsd.org/hardenedbsd/HardenedBSD.git synced 2024-11-14 14:21:18 +01:00
Code Issues Actions Packages Projects Releases Wiki Activity

libdiff: Detect and recover from file truncation.

Browse Source 
If a memory-mapped file is truncated before we get to the end, the
atomizer may catch SIGBUS.  Detect that, reduce the input length to
what we were actually able to read, and set a flag so the caller can
take further action (e.g. warn the user and / or start over).

Sponsored by:	Klara, Inc.
Reviewed by:	markj
Differential Revision:	https://reviews.freebsd.org/D45217
This commit is contained in:
Dag-Erling Smørgrav 2024-05-20 15:26:33 +02:00
parent 93fbfef0b5
commit 974ea6b297
2 changed files with 31 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
contrib/libdiff/include/diff_main.h
View File

@ -118,6 +118,7 @@ struct diff_data {
/* Flags set by file atomizer. */ /* Flags set by file atomizer. */
#define DIFF_ATOMIZER_FOUND_BINARY_DATA 0x00000001 #define DIFF_ATOMIZER_FOUND_BINARY_DATA 0x00000001
#define DIFF_ATOMIZER_FILE_TRUNCATED 0x00000002
/* Flags set by caller of diff_main(). */ /* Flags set by caller of diff_main(). */
#define DIFF_FLAG_IGNORE_WHITESPACE 0x00000001 #define DIFF_FLAG_IGNORE_WHITESPACE 0x00000001

35
contrib/libdiff/lib/diff_atomize_text.c
View File

@ -16,6 +16,8 @@
*/ */
#include <errno.h> #include <errno.h>
#include <setjmp.h>
#include <signal.h>
#include <stdbool.h> #include <stdbool.h>
#include <stdint.h> #include <stdint.h>
#include <stdio.h> #include <stdio.h>
@ -122,10 +124,18 @@ diff_data_atomize_text_lines_fd(struct diff_data *d)
return DIFF_RC_OK; return DIFF_RC_OK;
} }
static sigjmp_buf diff_data_signal_env;
static void
diff_data_signal_handler(int sig)
{
siglongjmp(diff_data_signal_env, sig);
}
static int static int
diff_data_atomize_text_lines_mmap(struct diff_data *d) diff_data_atomize_text_lines_mmap(struct diff_data *d)
{ {
const uint8_t *pos = d->data; struct sigaction act, oact;
const uint8_t *volatile pos = d->data;
const uint8_t *end = pos + d->len; const uint8_t *end = pos + d->len;
bool ignore_whitespace = (d->diff_flags & DIFF_FLAG_IGNORE_WHITESPACE); bool ignore_whitespace = (d->diff_flags & DIFF_FLAG_IGNORE_WHITESPACE);
bool embedded_nul = false; bool embedded_nul = false;
@ -136,8 +146,22 @@ diff_data_atomize_text_lines_mmap(struct diff_data *d)
ARRAYLIST_INIT(d->atoms, 1 << pow2); ARRAYLIST_INIT(d->atoms, 1 << pow2);
sigemptyset(&act.sa_mask);
act.sa_flags = 0;
act.sa_handler = diff_data_signal_handler;
sigaction(SIGBUS, &act, &oact);
if (sigsetjmp(diff_data_signal_env, 0) > 0) {
/*
* The file was truncated while we were reading it. Set
* the end pointer to the beginning of the line we were
* trying to read, adjust the file length, and set a flag.
*/
end = pos;
d->len = end - d->data;
d->atomizer_flags |= DIFF_ATOMIZER_FILE_TRUNCATED;
}
while (pos < end) { while (pos < end) {
const uint8_t *line_end = pos; const uint8_t *line_start = pos, *line_end = pos;
unsigned int hash = 0; unsigned int hash = 0;
while (line_end < end && *line_end != '\r' && *line_end != '\n') { while (line_end < end && *line_end != '\r' && *line_end != '\n') {
@ -164,15 +188,16 @@ diff_data_atomize_text_lines_mmap(struct diff_data *d)
*atom = (struct diff_atom){ *atom = (struct diff_atom){
.root = d, .root = d,
.pos = (off_t)(pos - d->data), .pos = (off_t)(line_start - d->data),
.at = pos, .at = line_start,
.len = line_end - pos, .len = line_end - line_start,
.hash = hash, .hash = hash,
}; };
/* Starting point for next line: */ /* Starting point for next line: */
pos = line_end; pos = line_end;
} }
sigaction(SIGBUS, &oact, NULL);
/* File are considered binary if they contain embedded '\0' bytes. */ /* File are considered binary if they contain embedded '\0' bytes. */
if (embedded_nul) if (embedded_nul)