libziparchive: ensure ReadAtOffset is atomic
ag/880725 modified ReadAtOffset to seek then read from the open file descriptor. Previously pread64 was used to provide atomic behaviour. This causes races when multiple threads are trying to access data from the file. This is supported, so this change reverts the relevant parts of the above CL to restore the old behaviour. Bug:27563413 Change-Id: I7bffd78da8c558745dfc3c072ba9691b1b15bb5b
This commit is contained in:
parent
dcdf300a1f
commit
b1911401e0
|
@ -152,6 +152,9 @@ void CloseArchive(ZipArchiveHandle handle);
|
|||
* if this file entry contains a data descriptor footer. To verify crc32s
|
||||
* and length, a call to VerifyCrcAndLengths must be made after entry data
|
||||
* has been processed.
|
||||
*
|
||||
* On non-Windows platforms this method does not modify internal state and
|
||||
* can be called concurrently.
|
||||
*/
|
||||
int32_t FindEntry(const ZipArchiveHandle handle, const ZipString& entryName,
|
||||
ZipEntry* data);
|
||||
|
|
|
@ -495,14 +495,20 @@ static int32_t UpdateEntryFromDataDescriptor(int fd,
|
|||
}
|
||||
|
||||
// Attempts to read |len| bytes into |buf| at offset |off|.
|
||||
// Callers should not rely on the |fd| offset being incremented
|
||||
// as a side effect of this call.
|
||||
// On non-Windows platforms, callers are guaranteed that the |fd|
|
||||
// offset is unchanged and there is no side effect to this call.
|
||||
//
|
||||
// On Windows platforms this is not thread-safe.
|
||||
static inline bool ReadAtOffset(int fd, uint8_t* buf, size_t len, off64_t off) {
|
||||
#if !defined(_WIN32)
|
||||
return TEMP_FAILURE_RETRY(pread64(fd, buf, len, off));
|
||||
#else
|
||||
if (lseek64(fd, off, SEEK_SET) != off) {
|
||||
ALOGW("Zip: failed seek to offset %" PRId64, off);
|
||||
return false;
|
||||
}
|
||||
return android::base::ReadFully(fd, buf, len);
|
||||
#endif
|
||||
}
|
||||
|
||||
static int32_t FindEntry(const ZipArchive* archive, const int ent,
|
||||
|
|
Loading…
Reference in New Issue