summaryrefslogtreecommitdiffstats
path: root/src
AgeCommit message (Collapse)AuthorFilesLines
2012-07-26recovery: removed commented-out declarations from header fileGravatar Nikias Bassen1-3/+0
2012-07-26restore: change UUID to UDID in error messageGravatar Nikias Bassen1-2/+2
2012-07-17restore: only print 'Attempt %d to connect...' in debug modeGravatar Nikias Bassen1-1/+1
2012-07-17tss: silence compiler warningGravatar Nikias Bassen1-1/+1
2012-07-17tss: Use https to contact webservices first as we sometimes hit a 404 over httpGravatar Martin Szulecki1-2/+6
After a while, Apple's servers have been begun redirecting to a 404 page using a 302 security redirect HTTP status code. By using a secure HTTPS connection retrieving TSS requests started to work fine again.
2012-07-17tss: Cycle through different hosts for TSS requests as a fallbackGravatar Martin Szulecki1-1/+9
It appears the number of HTTP requests from one IP to the TSS signing servers is limited by each signing host. This workaround increases the volume of devices that can be processed due to falling back to another signing host in case request limiting is in effect by the original host.
2012-07-17tss: Use appropriate @HostPlatformInfo value in TSS requestsGravatar Nikias Bassen1-2/+14
2012-07-17restore: Format TSS baseband request message strings correctlyGravatar Martin Szulecki1-2/+2
2012-07-17tss: Don't crash on errors like failed connection attempts and be more verboseGravatar Martin Szulecki2-7/+18
2012-07-17main: use correct filename after extracting into cache dirGravatar Nikias Bassen1-0/+2
2012-07-17common: silence warning by adding missing includeGravatar Nikias Bassen1-0/+1
2012-07-17main: implement dynamic filesystem extraction (into cache dir)Gravatar Nikias Bassen1-5/+36
2012-07-17main: return -2 instead of -1 if nonce could not be retrievedGravatar Martin Szulecki1-1/+1
2012-07-17main: refactor filesystem extraction and use cache folderGravatar Nikias Bassen1-61/+59
2012-07-17main: add --cache-path parameter and use it for version.xml and wtf imageGravatar Nikias Bassen3-14/+72
2012-07-17main: remove file locking as it appears to hang on some platformsGravatar Martin Szulecki1-9/+0
2012-07-17dfu: make sure client member of struct is NULLed correctlyGravatar Nikias Bassen1-1/+8
2012-07-17restore: handle -1 errorGravatar Nikias Bassen1-0/+6
2012-07-17main: do not delete cached filesystemGravatar Nikias Bassen1-13/+16
2012-07-17main: use extracted filesystem if presentGravatar Nikias Bassen1-0/+24
2012-07-17ipsw: add function ipsw_get_file_size() to get uncompressed file sizeGravatar Nikias Bassen2-0/+28
2012-07-17tss: Do not output "MESSAGE=" prefix of TSS server error messagesGravatar Martin Szulecki1-1/+1
2012-07-17restore: Error out if writing final personalized bbfw archive file failedGravatar Martin Szulecki1-3/+6
2012-07-17restore: Always try to receive and handle possible StatusMsg on errorsGravatar Martin Szulecki1-7/+7
This does not directly exit if anything breaks but attempts to read one more message from restored which usually is a StatusMsg and contains information about the error that occoured.
2012-07-17restore: Handle printing logs of a StatusMsg if availableGravatar Martin Szulecki1-0/+12
2012-07-17asr: Output byte counts in error message if sending data failsGravatar Martin Szulecki1-1/+1
2012-07-17asr: Fix memory leak in asr_send()Gravatar Martin Szulecki1-4/+2
2012-07-17asr: Add macros for other values of validation requestGravatar Martin Szulecki1-4/+8
2012-07-17asr: implemented checksumming for filesystem uploadGravatar Nikias Bassen3-34/+130
2012-07-17mbn: Use correct log debug() function instead of printf for warningsGravatar Martin Szulecki1-1/+1
2012-07-17restore: Return -2 error code on ASR failures as a new attempt might workGravatar Martin Szulecki1-1/+1
2012-07-17restore: If opening a device times out completely, make error unrecoverableGravatar Martin Szulecki1-3/+4
2012-07-17restore: Don't process further device add events if we already have a matchGravatar Martin Szulecki1-1/+1
2012-07-17restore: Print message if a device is rebooted in restore modeGravatar Martin Szulecki1-0/+1
2012-07-17recovery: Send "preventive" ZLP directly after sending kernelcache componentGravatar Martin Szulecki1-2/+2
2012-07-17main: If mode switching from normal to recovery fails, return error code -5Gravatar Martin Szulecki1-1/+1
This should indicate that a manual mode switch is required.
2012-07-17recovery: Make sure a connection is available before sending iBEC, tooGravatar Martin Szulecki1-0/+6
The code path without TSS request did not initialize the recovery client in order for it to be used to send iBEC. Now the client is always initialized.
2012-07-17main: Report device state again after a restore rebootGravatar Martin Szulecki1-0/+1
2012-07-17restore: refactor device detection to find device by ECIDGravatar Nikias Bassen1-28/+88
2012-07-17restore: Normalize messages printed in handlers of main restore process loopGravatar Martin Szulecki1-4/+17
2012-07-17main: Do not wait 15 seconds for device to enter restore as we use events nowGravatar Martin Szulecki1-3/+0
2012-07-17restore: Correctly print success message after receiving baseband SHSH blobsGravatar Martin Szulecki1-0/+1
2012-07-17restore: Remove commented obsolete code calling idevice_unsubscribe()Gravatar Martin Szulecki1-2/+0
2012-07-17restore: Make restore mode device connection handling more verboseGravatar Martin Szulecki1-1/+3
2012-07-17restore: Only print "Provisioning:" if IMEI is available at allGravatar Martin Szulecki1-1/+1
2012-07-17tss: Print server MESSAGE response if unhandled error is returnedGravatar Martin Szulecki1-1/+2
2012-07-17restore: Only print plist if errors occoured or in debug modeGravatar Martin Szulecki1-4/+6
2012-07-17dfu: Do not send another ZLP after sending iBEC as it breaks DFU mode switchGravatar Martin Szulecki1-1/+1
2012-07-17restore: use idevice event to wait for device entering restore modeGravatar Nikias Bassen1-28/+20
2012-07-17USB: Send a ZLP after mode switches to hopefully increase stabilityGravatar Martin Szulecki2-0/+7
Sometimes devices do not switch modes, sending a zero length packet appears to fix this situation and prevents failures during restore.