HDDOSD Issues

I am having an issue, I have a fat ps2, SCPH-36001, and have been using the 48 bit HDDOSD image made by vash32. I was using the latest version of ulaunchelf_khn to update the version of popstarter that I had on the drive to the newest version, and it failed to write to the partition and then showed the dir as empty afterwards.

I restarted the ps2, and then it started to complain that the HDD information had been corrupted, so I pulled the hard drive out and plugged it into my computer and reimaged the drive. The ps2 now refuses to detect the HDD, it spins up, sounds like its writing to it, and then the hdd spins down and the ps2 goes to the rom-based osd

Could anybody help me with this?

 

Re-write the HDDOSD image to your PS2HDD;
Do a scan/repair with WinHIIP;
Check the HDD with HDDChecker.

 

I have rewrote the image multiple times now, Scan and repair shows no damage to repair but I save it anyway to make sure, and I will have to get back to you on the checking with HDDChecker, as my only way to run homebrew on the fat ps2 was the HDD, will have to dig out my slimline and install FMCB or something. Is there anything I can do PC-side to check the hard drive is created correctly from the image, and is there any way for me to make sure it's not my network adapter going bad.

If it helps any, the MAC address of the adapter does not show up in the system information, although I dont remember if it does in Browser version 1.1.6

Edit - The hard drive does not boot up Browser 2.00 on my 39001 either, I just checked it to make sure DEV9 didnt die or something stupid

 

Just tried another entire hard drive, same result. Drive spins up, looks at something, then ps2 turns interface off and boots to rom osd

Also tried a different image, same result. Is there any way to verify that the network adapter is working correctly?

 

Run an integrity check on the drive to insure it doesn't have corrupted sectors.

 

I already did, as well as changing drives completely, no change

 

It's a little hard to say what's wrong because you used a rather hack'ish way to get the HDD Browser update onto your HDD.
If you can get FMCB installed for this console, it would be great. You can then try to format the HDD with uLaunchELF, before installing FHDB onto the HDD. If FHDB can be installed properly and it works, then you know for sure that the HDD (and hence the network adaptor) is fine.

 

The only reason I used the image is because it's currently my only way to run homebrew. I had a slimline that was modded to do the disc swap trick, but it is currently in storage. Is there any other way than the disc swap trick for the Fat PS2?

 

You could use this to format the HDD and to install FHDB, although it is an older version: http://psx-scene.com/forums/f153/un...e-package-obt-119562/index13.html#post1141776

 

I will try this when I get back to my house, found one of my other network adapters as well so I will try that as well. Has there ever been mentions of DEV9 ports ever going bad, or am I safe in that regard.

 

Yes I've seen them die before on my friend once and he never used HDDOSD or my img before, just Open PS2 Loader.
It maybe have died or some parts or caps maybe? Don't know... Anyway I've just made a newer img that using sp193 HDDOSD 1.00J 48BIT that not as hacked up & may not be as buggy. HDDOSD 1.00J LBA48 (48BIT)

 

Awesome! I will try it when I get a chance. Will a J update work on a USA console? I though the updates were region locked. Also, quick question, is there any way to get around the issue where the Browser slows down tremendously after a certain point or is that just a side effect of forcing a 24bit program to read 48bit

 

That'll be great.

It's not quite the expansion interface going bad... but usually it's the console's ability to supply power properly to the HDD unit.

The files from me are region-free, although the browser was from the v1.00J disc.

Unless I am wrong, I believe that it's caused by the poor design of APA: insufficient caching system and no allocation table.

On a small disk (i.e. the official 40GB unit), there is no problem because the APA driver (HDD.IRX) will be initialized by the HDD Browser with sufficient buffers, to cache access to all partitions on the disk. However, when the number of partitions on the disk increases past a certain point (possible on sufficiently large disks), the number of buffers become insufficient.

Unlike a filesystem like FAT or EXT2, there is no allocation table of any sort. So to locate a partition, the disk has to be scanned for the desired partition. There is a cache system to record the information of the partitions on the disk, but its capacity is limited by the programmer (and must fit within IOP RAM).

When the browser is listing the partitions (game installations) on a disk, it probably opens the partitions displayed, to retrieve their icons.
Due to the lack of an allocation table, the partition to be opened (i.e. to retrieve the icon) has to be scanned for. If there are insufficient buffers, the APA driver starts slowing down because it will start reading through all uncached partitions on the disk, after a certain point.
So if there are 200 partitions on your disk and only 126 buffers, the remaining 74 partitions will have to be scanned for. This is done for every single open operation, after the 126th partition.

It doesn't seem like much to read, but part of the problem is that this is APA. The partitions are spread out across the disk, so the heads will also spend time seeking.

 

Thanks for the explanation, I figured it was something to do with caching, just wasnt sure.

 

Tried the image with a new network adapter, worked first time. I am having an issue with WinHIIP though, the image is detected as toxicOS, and when I switch it to HDLoader 48-bit it offers to make the HDLoader partition and then freezes if I put yes

 

This only happens when I write the image, if I use winhiip to format the drive to toxicos and then convert it to hdloader 48-bit the partition creates fine

 

I never tested it on WinHIIP however I've used p193's HDLGameInstaller & HDL Dump GUI 1.5 no problem, I think the newest var. of hdl_dumb & hdl_dump work as well.

I'm using a 320GB 7200rpm HDD & over 70 PS1/PS2 games on it & only 140gb free.. no slow downs so far.. maybe once I get over 126 partitions... However what I fine funny is that this new setup I've made boots into HDDOSD almost as fast as booting a PS2 using no HDD on my SCPH-18000 & SCPH-39001.

Any way of maybe fixing this bug or speeding up the scanning.. by say using a SATA or SSD? The IDE speed is at like Ultra ATA 66 or something like that?

 

I just tested another image, HDDOSD_100J_ATADP_DTL-H_FIX_48BIT, which I believe is the same image just without the added apps, and it does not exhibit the issue. Finally narrowed it down to the program freezing anytime it touched the partition records after writing the image you told me to use

 

Yes its the same image, just made before adding the apps to it. Well maybe its from it not having a "__net" partition or the other partitions only being 128MBs? I can't test WinHIIP right now myself but I've tested HDL Dump Helper GUI 2.3 before and it works fine on it that image & can to network as well. (P.S. Have you tested your HDD to see if its bad any? I full tested my HDD before making this image & I zero filled the HDD)

As for HDDOSD slow downs, maybe a way to edit how it shows partitions as it list the newest 1st... if you make the icons that slow down go at the end, it will only slow down at the end of the list and not at the start. That does not fix the problem but is a lot nicer.

Anyone try using a "IDE to SATA" & a SSD on HDDOSD? I have a one and thinking of try a SATA 500GB to see if its any faster on the slower parts but I'm thinking a SSD or SSHD will work out better?

 

I do not believe that the drive is bad, as I am only having the issue with that specific image. With the other image, I have been able to add 50+ games with WinHIIP to the drive without a issue. It almost seems like the program is choking on something to do with whatever is in the image after the useful data. The reason I think this is because the program can see the partitions with the scan & repair and can even repair the drive perfectly fine. Its when the program tries to add any data that it dies.

The only time I have a issue with the HDDOSD slowing down is when I have a certain amount of partitions already created on the disk, and it only slows down when I go down to the bottom of the list, which causes it to have to open the partitions to grab the info for the games.

What is the max speed that the PS2 supports? I doubt that a SSHD would help much as its SSD cache would constantly be in flux due to the PS2 not doing any really repeated reads other than the OSD and system files if im not mistaken. A SSD would probably help, if nothing else to cut down on seek times while reading the partitions initially. Assuming that the adapter converts the commands correctly and effeciently, I dont think there would be that much of a difference between PATA and SATA really