Age | Commit message (Collapse) | Author | Files | Lines | |
---|---|---|---|---|---|
2024-09-21 | asr: Fix memory corruption due to double free | 1 | -2/+2 | ||
2024-09-21 | asr: Add support for second Initiate request | 1 | -16/+43 | ||
First observed in iBridgeOS 9.0. The first Initiate ASR packet (checksum_chunks = false) requests 64 bytes of the IPSW at offset 0, after which another Initiate follows requesting a switch to (checksum_chunks = true) and additional OOBData. | |||||
2024-09-19 | Also print libirecovery version | 1 | -2/+2 | ||
2024-09-19 | Replace sprintf with snprintf | 4 | -28/+30 | ||
2024-09-19 | restore: Add SupportedAsyncDataTypes for both iOS and macOS | 1 | -7/+9 | ||
2024-09-18 | Print libtatsu version alongside idevicerestore version | 1 | -2/+2 | ||
2024-09-18 | Only print libimobiledevice debug info for debug level > 2 | 1 | -1/+3 | ||
2024-09-18 | common: Try to improve terminal output with fflush() | 1 | -0/+3 | ||
2024-09-18 | restore: Fix incorrect fallback case | 1 | -2/+2 | ||
When the updated behavior is not triggered, the legacy behavior must be correctly executed. Thus, always correctly fall back to old behavior instead of aborting here. For example message can be NULL when restoring iOS 1.0 (in my fork). | |||||
2024-09-18 | restore: Only send FirmwareData when it has been requested | 1 | -21/+101 | ||
2024-09-18 | restore: Make sure to error out when async data request handler can't be started | 1 | -0/+3 | ||
When passing --ignore-errors, it would not fail when the async data request handler thread cannot be started, and might end up waiting forever for something to happen. | |||||
2024-09-18 | restore: Make wait for URLAsset on first chunk optional | 1 | -5/+15 | ||
2024-07-09 | ipsw: Fix concurrent access to ZIP file | 2 | -35/+137 | ||
With the new AsyncDataRequestMsg we have the problem that multiple threads would access the .ipsw, however we were only using one struct zip* handle for the entire procedure, resulting in read errors when concurrent access occurs. This commit fixes it by opening the zip file for every access separately. | |||||
2024-07-01 | Fix heap buffer overflow in URLAsset handling | 1 | -1/+1 | ||
2024-06-29 | Remove debug printf | 1 | -1/+0 | ||
2024-06-26 | restore: Always try to use DeviceGeneratedRequest data for TSS requests if ↵ | 1 | -15/+45 | ||
present | |||||
2024-06-24 | Add support for iOS 18 restore process | 7 | -131/+750 | ||
2024-06-22 | restore: Remove `device` from the parameters of the functions | 2 | -19/+29 | ||
2024-06-22 | restore: Remove `build_identity` from the parameters of the functions | 2 | -100/+178 | ||
2024-06-16 | Remove OpenSSL dependency in favor of libimobiledevice-glue's hash functions | 9 | -800/+12 | ||
This also removes the sha1/sha384 code from this repository since we are using the ones from libimobiledevice-glue now. | |||||
2024-06-12 | Add missing linebreak to log message | 1 | -1/+1 | ||
2024-06-05 | Small change to align with updated libtatsu | 1 | -1/+1 | ||
2024-05-23 | Link against the new libtatsu and remove tss code | 8 | -1876/+11 | ||
2024-05-18 | Add missing cast to silence compiler warning | 1 | -1/+1 | ||
2024-05-14 | Require libplist 2.6.0 | 5 | -71/+56 | ||
2024-05-05 | Updated to use libplist 2.5.0 API | 8 | -431/+284 | ||
2024-04-04 | dfu: A little code optimization | 1 | -15/+5 | ||
2024-04-04 | Remove annoying linebreak | 1 | -2/+1 | ||
2024-04-03 | Fix restore mode component personalisation | 1 | -23/+31 | ||
2024-03-23 | Add support for Port DFU device restore | 8 | -0/+509 | ||
2024-03-23 | tss: Update libauthinstall verison string | 1 | -1/+1 | ||
2024-03-23 | Fix some variable types for more consistency | 8 | -24/+24 | ||
2024-03-08 | Make sure to extract the build manifest before doing restore mode checks | 1 | -14/+15 | ||
Otherwise we could set tss_enabled to 0 but the extraction would re-enable it. | |||||
2023-11-27 | restore: Only print progress bar for images larger than 16 MB | 1 | -1/+1 | ||
2023-11-21 | Update libzip API usage to use non-deprecated functions | 1 | -11/+12 | ||
2023-11-15 | Change path_get_basename arg to const too | 2 | -2/+2 | ||
2023-11-15 | Change path_get_basename()'s return type to const char* | 3 | -6/+5 | ||
This makes it clear that the return value is immutable and moreover suggests that the return vale is not allocated and thus should be treated carefully. | |||||
2023-11-15 | restore: Fix UaF | 1 | -2/+2 | ||
`fsname_base` points inside the dynamically allocated `path` which is freed before `fsname_base` is used, creating a use-after-free condition. This commits makes sure to free `path` only after it is no longer needed. | |||||
2023-11-15 | asr: Fix sending payload without checksum | 1 | -4/+6 | ||
2023-11-09 | Fix update restore by making sure the premanifest is properly generated | 3 | -14/+32 | ||
2023-11-09 | Print progress for large components (e.g. Cryptex) | 5 | -13/+41 | ||
2023-11-07 | Print device Product and Build Version and IPSW Product and Build Version | 2 | -18/+26 | ||
It wasn't entirely clear what "Product Version" and "Product Build" would be so prefixing it with "IPSW" makes it clear it's the version being restored. | |||||
2023-11-02 | Extract OS component when using older ipsw archives | 7 | -11/+143 | ||
Older ipsw archives have the root filesystem stored in compressed format rather than just "stored". The "Verifying Filesystem" step would then fail as compressed files are not seekable in ZIP files. This commit introduces a detection for this and has the filesystem extracted should it be required. If not using a cache path, the temp file used for extraction will be deleted after the procedure is completed. | |||||
2023-10-09 | tss: Add USBPortController1,* entries to parameters | 1 | -0/+4 | ||
2023-10-07 | Improve debug output by suppressing libimobiledevice and libirecovery debug ↵ | 2 | -3/+7 | ||
output by default To get libimobiledevice and libirecovery output, add -d or --debug twice. | |||||
2023-10-06 | restore: Also print checkpoint warning messages | 1 | -0/+4 | ||
2023-10-06 | restore: Add Ace3 as known updater name to suppress error message | 1 | -0/+6 | ||
2023-10-06 | restore: Skip adding FirmwareData to FirmwareResponseData for Rose | 1 | -0/+6 | ||
2023-10-04 | restore: Improve checkpoint log output again, make sure to always check for ↵ | 1 | -5/+8 | ||
errors Turns out even with a CHECKPOINT_RESULT of 0 we can still have a CHECKPOINT_ERROR string. | |||||
2023-10-04 | restore: Refine checkpoint log output | 1 | -6/+15 | ||