Tutorial
Writing the disc, running it, and reading what comes back. Every console firmware version matters here, so check yours first.
Each statement below carries a label. V means I measured it from the published files and you can re-measure it. I means I inferred it from the disc's own structures and the strings its code prints. P means ordinary disc-burning practice, which I did not measure.
I ran this on my own PS4 at 14.00 and it worked, so the flow below is what I expect you to meet. The exact text the disc prints is quoted from the payload's class files, so it is word-perfect. Your burner, disc brand and console revision are different, and the result is yours to confirm at your own risk.
Contents
- Pick the image
- Verify the download
- Look inside without burning
- Write the disc
- Prepare the console
- Run it
- Every message the disc prints
- Confirm the result
- When it fails
- Troubleshooting
Pick the image
Burn one file and keep its hash. hen.iso is the canonical disc and matches the reference digest for the Bonk HEN 14.00 release. hen-branded.iso carries an updated credit line in NOTICE.txt and behaves the same.
Verify the download
# Windows PowerShell
Get-FileHash -Algorithm SHA256 .\hen.iso
# Linux and macOS
sha256sum hen.iso
Compare against the table on the home page, or run sha256sum -c SHA256SUMS.txt in the folder you downloaded everything to. A corrupted download stops at preflight with PREFLIGHT FAIL: and does nothing else, which is the safe failure, but it costs you a boot.
Look inside without burning
Mount the image or open it in 7-Zip. You should see these sizes, all verifiable [V]:
BDMV/BDJO/00000.bdjo 164 B auto-run applet: org.bdj.InitXlet
BDMV/JAR/00000.jar 192,403 B signed payload, 78 entries
BDMV/STREAM/00000.m2ts 3,975,168 B the stream the player opens
BDMV/PLAYLIST/00000.mpls 170 B
BDMV/CLIPINF/00000.clpi 1,136 B
BDMV/index.bdmv 526 B
BDMV/MovieObject.bdmv 7,402 B
BDMV/AUXDATA/00000.otf 234,304 B BD-J fonts
NOTICE.txt 564 B safety notice, sector 448
x/offline/Hx1400.class 56,859 B kernel stage, loaded from the disc
CERTIFICATE/id.bdmv, app.discroot.crt, bu.discroot.crt
UDF 2.50, volume label BONKHEN, 3,040 sectors of 2,048 bytes
If your tool cannot read the volume label, try a different tool before deciding the image is wrong. UDF 2.50 with UCS-2 directory records trips up some older extractors.
Write the disc
The image has to go on as a raw sector image. Copying hen.iso onto a disc as a file gives you a data disc with nothing to auto-play, and that is the single most common reason a disc "does nothing".
Windows, ImgBurn P
- Write image file to disc, select your ISO.
- Pick the slowest write speed the drive and disc both support.
- Tick Verify.
- Write, then let the session close. Leave it single-session and finalized.
Windows, AnyBurn P
Copy/Write image file to disc, with the same speed and verify choices.
Linux P
sudo growisofs -dvd-compat -speed=2 -Z /dev/sr0=hen.iso
isoinfo -d -i /dev/sr0 # expect: Volume id is "BONKHEN"
macOS P
hdiutil burn hen.iso
Prove the payload survived the burn
The JAR sits at sectors 596 through 689. Reading it back off the disc and matching its hash rules out a bad write in one step [V]:
# Linux
dd if=/dev/sr0 bs=2048 skip=596 count=95 2>/dev/null | dd bs=1 count=192403 2>/dev/null | sha256sum
# Windows: mount the disc, then
(Get-FileHash E:\BDMV\JAR\00000.jar -Algorithm SHA256).Hash
Both must print d6b16769366df24d50623afdaefab6b2299ff02bb816ea30da7839a572b3c0ed. If they do not, re-burn slower or change disc brand before you blame the exploit.
Prepare the console
- Check firmware under Settings, System, System Information. Anything other than 14.00 means this disc is the wrong tool.
- Copy important saves to USB or cloud. A panic during a write is the realistic data-loss path.
- Take the console offline for early runs. Unplug Ethernet or cut Wi-Fi at the router. Safety explains what that buys you.
- Shut the console down completely, so you start from a clean boot. Rest mode does not count.
- Set a phone camera to record the screen. Reading a report line after a reboot beats remembering it.
Run it
- Insert the disc on the home screen while signed into a normal user profile.
- The player opens the disc and the BD-J applet starts on its own [V]. You do not choose it from a menu.
- A HAVI screen appears headed
BONK HEN 14.00with a running log. The twenty-five preflight checks happen in a background thread. Wait for them. - When
LOCAL PREFLIGHT PASS. Press X once to run.appears, press X exactly once. - Let the report lines finish. Once
Entering one-shot Hx/HEN stageis on screen, the kernel stage is live, so leave the disc, the controller and the power alone. - Keep the disc inserted until you have read the last line and recorded the outcome.
Every message the disc prints
Quoted from the payload's class files [V].
Preflight, healthy
| On screen | Meaning |
|---|---|
Bonk HEN local asset preflight | Preflight started. |
Four local helper classes verified | The four x/offline classes match their pins. |
All 40 local command templates verified | All forty cmd-*.bin templates match their pins. |
Exact HEN image verified in signed JAR | bonk/hen.bin matches its pin and came from the signed JAR. |
Command 32 uses verified local HEN URL | The HEN path resolved to the local file: resource. |
Command 2 helper status: … | Helper object construction report. |
Native code-window branch verified | Code-window expectations hold. |
LOCAL PREFLIGHT PASS. Press X once to run. | Your cue. Press X once. |
Preflight, stopped
| On screen | Meaning and what to do |
|---|---|
PREFLIGHT FAIL: reason | An asset missed its pin and nothing executed. Re-verify the download and re-burn. This fixes most cases. |
Xlet is not in a disc JAR | The payload thinks it is not running from a real disc JAR. Usually the ISO was copied as a file. Re-burn raw. |
HEN URL is not verified local JAR resource | The HEN image did not resolve to the signed JAR. Check media and burn. |
Hx diagnostic shape changed, HEN assignment framing changed, HEN sentinel not unique | Payload structure differs from what the verifier expects. That points at a corrupted read rather than a console fault. Re-verify and re-burn. |
Marker scan: no match (-1), Marker scan: helper exception (-2) | Marker plan scan found nothing or errored. Re-burn, and report it if it repeats. |
Run and result
| On screen | Meaning |
|---|---|
Attempt 1 started: local interpreter commands | The gated sequence began. There will be no attempt 2 in this boot. |
Bonk HEN one-shot local run | Executing the reconstructed templates. |
STOP at command N | A stage reported a result that did not match its expected shape. Nothing after N ran. Note N. |
ledger stopped | The result ledger ended the sequence. |
Entering one-shot Hx/HEN stage | Kernel stage running. Touch nothing. |
Hx summary: … | The kernel stage's own summary. |
Hx reports HEN start. Verify menu and kernel R/W separately. | Best-case marker, and the payload declines to call it success. Go to step 8. |
H8 kept: hen map | The HEN mapping survived the stage. |
H2 /data/hen audit: … | Audit of the /data/hen object. |
K NOT fully applied | Failure. The patch set is incomplete. |
Hx reported failed pairing | Failure. Kernel-object pairing did not complete, and the disc warns this stage may still have written unknown kernel objects. |
Hx reported incomplete kexec patch | Failure. The syscall path was not fully patched. |
Hx status remained pending | No verdict returned. |
Hx still running; read-only status … | Stage unfinished; status read without touching it. |
jitshm mmap for hen failed, helper setup failed, marker scan failed, local HEN command too long | Failure in the named step. |
No retry in this boot. Cold reboot before another attempt. | Follow step 9 exactly. |
Hx final report missing, Hx success summary incomplete, Hx HEN or patch log incomplete, Hx report could not be decoded for display | The result could not be read. Treat the run as unknown rather than successful. |
Confirm the result
The on-disc notice says a displayed success report needs independent verification on the console, and the payload repeats the instruction. A green line is the program's opinion about its own work. Check both halves:
- Kernel read/write. A homebrew loader has to observe patched kernel access. If you have no way to test that, you cannot claim HEN applied.
- The HEN menu. The native stage embeds a settings UI. Its own resources define these toggles, with the offsets inside the image [V]:
| Offset | Toggle | Its own description |
|---|---|---|
| 0x03b813 | Enable exploit fixes | Attempts to repatch bad or missing patches that should have been made by the exploit, to keep the system stable after the initial jailbreak. |
| 0x03b911 | Enable memory mapping patches | Memory mapping patches for decrypting binary files, so the system can handle encrypted executables and libraries. |
| 0x03ba0e | Block firmware updates | Unmounts the /update partition so an update cannot land. The disc asks you to confirm this one. |
| 0x03c3f8 | Write HEN plugin files to HDD | Writes the default HEN plugin PRX files to the internal drive. The image's own text recommends this for first-time setup. |
If that menu is reachable, the menu half is satisfied. If it is not reachable, HEN did not apply regardless of what the disc printed. On a first success, turn on the plugin write option and leave the other three off until you have a reason for them, since each one changes kernel or filesystem state.
Once you have a working run, send me a photo or screenshot of the screen. Firmware, model and the last line are the useful parts.
When it fails
- Do not retry inside the same boot. The disc enforces this and states it twice, and the native stage warns that armed-but-missed patches are cleared by a cold reboot. Retrying on top of leftover kernel state is the reliable way to turn a failed attempt into a corrupted one.
- Cold reboot properly: hold the power button until the console is fully off, or unplug it for about thirty seconds. Restart from the menu and rest mode both fail to clear it.
- Power on and make one attempt.
- Two failures with the same
STOP at command Nor the same preflight reason means stop and report it. The exact line and your firmware are the whole diagnosis.
Troubleshooting
| Symptom | Likely cause | Do |
|---|---|---|
| Console says the disc is unsupported | ISO copied onto the disc as a file | Re-burn in raw image mode |
| Disc reads but nothing auto-plays | Player opened a data or movie view | Re-burn slower, finalize, try another disc brand |
| Black screen, no text | Drive read errors, or BD-J fonts not rendering | Re-check the JAR read-back from step 4 |
PREFLIGHT FAIL on a hash line | Corrupt burn or corrupt download | Check the download hash, then re-burn |
Freeze at Entering one-shot Hx/HEN stage | Kernel stage is live | Wait it out, then follow step 9 |
| Panic or boot loop after a reported success | The stage wrote further than expected | Stop. Cold reboot. Console health comes before the experiment. |
| Worked once, never again after an update | Firmware moved off 14.00 | Nothing this disc can do about it |
Reporting a problem
Send these in order: firmware exactly as the console prints it, which image you burned and its verified hash, burner model, disc brand, write speed, whether verify-after-write passed, and the exact last line on screen. Say clearly whether LOCAL PREFLIGHT PASS appeared at all, because that one fact separates an asset mismatch from a declined kernel stage, and they are different problems. Keep serials, MAC addresses and PSN credentials out of it.