Does MTD/Yaffs2 perform better than BML/Rfs,Ext4? - Epic 4G General

By now maybe there are some available performance results, especially since it's possible to run the same rom with different file systems.
I haven't seen much discussion about the performance aspects even in theory.
Is yaffs2 faster? Is it more stable or less likely to corrupt files? Does it result in more disk writes? Is it harder on the storage medium?
So beyond the fact that MTD/YAFFS2 is more mainstream and will make CM easier to keep up with what, if any are the performance benefits of the filesystem itself? Especially for people who may never go to CM or won't for months.
Oh, last question: MTD allows more storage in the /data folder if I understand correctly. How much more though? What is the /data size under BML and how large can it get under MTD? And what performance benefit is there from that?

More stable, faster, less writes, easier on disk, open source, dev customizable for partition sizes (not just data)...
Sent from my SPH-D700 using XDA App

I switched over to MTD yesterday and so far my phone is running fine. The switch took half an hour and I notice a little less lag with
certain apps.
IMO, you need to try it for yourself and see if you like it.
Send from my SPH-D700 (Samsung Epic 4G)

garyf said:
By now maybe there are some available performance results, especially since it's possible to run the same rom with different file systems.
I haven't seen much discussion about the performance aspects even in theory.
Is yaffs2 faster? Is it more stable or less likely to corrupt files? Does it result in more disk writes? Is it harder on the storage medium?
So beyond the fact that MTD/YAFFS2 is more mainstream and will make CM easier to keep up with what, if any are the performance benefits of the filesystem itself? Especially for people who may never go to CM or won't for months.
Oh, last question: MTD allows more storage in the /data folder if I understand correctly. How much more though? What is the /data size under BML and how large can it get under MTD? And what performance benefit is there from that?
Click to expand...
Click to collapse
1. From my own personal experiences, yes it contains a quicker response in areas in which it is needed, more of a smooth user experience. I've not yet seen someone complain about a clunkiness due to our new partition configuration. Yes it is more than stable, data corruption on MTD/yaffs2 is unheard of, very unlikely as it doesn't just randomly occur or nowhere as often as it can on EXT4, it doesn't mean that if you do something like OC to a high frequency that it cannot happen. That's always is and always will be subject to how you run your device. And no, it is more meant for flash based devices which we in essence are. It's a ways smoother and reads and writes better to our nand.
2. To a normal user, in all honesty, it's something new. The benefits are the increase in space of data. 689 mb to be exact, although some of that space is always being used from the getgo. Being that there are now stock boot.imgs any user can go to MTD and not truly have a need to go back to BML or the constant fear of having to Odin for those who are iffy on it and not lose out on the stock feel, they've a choice between AOSP as well as StockWiz.
3. 689 mb MTD compared to about 200ish BML, or less in the data. And you've got to take space from the others to allocate it in the longer run of things. It doesn't stunt or skyrocket performance, it's independent of it.
Let me know if I answered your questions at all lol. I was vague as usual.

I've battery pulled dozens of times on yaffs2 with no data corruption. no such luck on ext4
Sent from my SPH-D700 using xda premium

I've not noticed one iota of difference between RFS, ext4 and now yaffs2 in real world usage, so all I can go on are the benchmarks I've run.
In my testing synthetic disk benchmarks have shown yaffs2 to perform about the same as RFS. ext4 is only faster when mounted with the unsafe, data-corrupting 'noauto_da_alloc' option. Without that options, it performs about the the same as RFS too.
yaffs2 is a win for all of the reasons marcusant listed aside from (IMO) performance.
---------- Post added at 08:57 PM ---------- Previous post was at 08:52 PM ----------
starskyrob said:
I've battery pulled dozens of times on yaffs2 with no data corruption. no such luck on ext4
Click to expand...
Click to collapse
You can thank the 'noauto_da_alloc' option for that. In all of the documentation I've found regarding ext4, it explicitly warned that if you use that option you must have battery backup for your disk devices or you will experience zero length files in the case of power loss.
Whoever first discovered that tweak on ext4 and convinced everyone to use should be smacked over the head a few times.

I think team bonsai disabled journaling
..
Sent from my SPH-D700 using XDA App

marcusant said:
I think team bonsai disabled journaling
..
Sent from my SPH-D700 using XDA App
Click to expand...
Click to collapse
When I was first making my kernel I noticed that my ROM running ext4 would only score 1700 in Quadrant while other ext4 ROMs scored 2300. It turned out to be the 'noauto_da_alloc' option that provided that instant 700 point boost. I also experimented with jourmaling and found the difference it made very small - at least in quadrant scores. The 'noauto_da_alloc' option leads to zero length files on power loss, regardless of whether or not journaling is on.
Oh, and I think I found the person who introduced this to everyone. It was supercurio.

I have to say I find this thread extremely informative.
Because of rfs/ext4, I expect there to be some form of tradeoff going between file systems. I'm a lot less worried about that now.
Thanks a lot all.
My understanding from the answers I've seen is that quadrant scores are about the same between rfs and yaffs2 and there are almost no real-world performance issues either way, except that it will take longer before you have to begin moving apps to sd storage.
If that's right, thanks for making it very easy to get that info.

marcusant said:
I think team bonsai disabled journaling
..
Sent from my SPH-D700 using XDA App
Click to expand...
Click to collapse
Ouch!...LOL! And as quickly as I could re-enable journaling I did. Did not notice any performance difference. Now we have MTD!

Another thing, the extra space on the /data partition only comes for roms built for MTD, correct?
Until TW roms are built from the ground up with MTD, rather than porting them across, that benefit will not apply?
Also can someone explain in more detail where this extra memory is coming from? Do BML filesystems just waste 400 megs or so of storage space, or is something being compressed or removed to get this extra storage?

garyf said:
Another thing, the extra space on the /data partition only comes for roms built for MTD, correct?
Until TW roms are built from the ground up with MTD, rather than porting them across, that benefit will not apply?
Also can someone explain in more detail where this extra memory is coming from? Do BML filesystems just waste 400 megs or so of storage space, or is something being compressed or removed to get this extra storage?
Click to expand...
Click to collapse
This extra space is on every MTD ROM out. The space won't change whether the ROM is intentionaly built on MTD or not. The extra space on BML is taken by /cache which is reduced to 25MB or so. /cache is really only used by market(for large app downloads) and by OTA updates(which we'll only see 1 more).
Sent from my SPH-D700 using xda premium

ac16313 said:
This extra space is on every MTD ROM out
Click to expand...
Click to collapse
WOW! I went and checked, and I have 267mb used and 409mb free now!
....and thats with NO apps on SD card!!!
Dammmmmmmmmmmm!
edit: 193mb used with 483mb free after moving apps to SD WOOHOO!
ac16313 said:
The extra space on BML is taken by /cache which is reduced to 25MB or so. /cache is really only used by market(for large app downloads)
Click to expand...
Click to collapse
Would this affect extra large app downloads, as in when the app is 50+mb or some such?

DCRocks said:
Would this affect extra large app downloads, as in when the app is 50+mb or some such?
Click to expand...
Click to collapse
Not if the market is fixed to cache to sdcard
Sent from my SPH-D700 using xda premium

ac16313 said:
Not if the market is fixed to cache to sdcard
Sent from my SPH-D700 using xda premium
Click to expand...
Click to collapse
thanks, gonna have to keep an eye on that the next time I dl a large app

ac16313 said:
Not if the market is fixed to cache to sdcard
Sent from my SPH-D700 using xda premium
Click to expand...
Click to collapse
There is an init script in my rom and cm7 that downloads large apps to data.
Sent from my SPH-D700 using XDA App

marcusant said:
There is an init script in my rom and cm7 that downloads large apps to data.
Sent from my SPH-D700 using XDA App
Click to expand...
Click to collapse
Can you make a flashable for the rest of us ROM hoppers?
Sent from my SPH-D700 using XDA App

Related

How much internal memory do you have left?

I have voodoo (dirrks) non overclocked, don't have a ton of apps installed but have just over 700mb free internally. I know its because of the voodoo partition but is there some way to relocate some things? Pretty sure my browser cache is set to external...
Feel like I'm gonna run out soon
Sent from my SCH-I500 using XDA App
i have 1.3 gigs
Voodoo should not take up a significant amount of extra space. The old ext2 lagfix did because it basically created a virtual partition within a partition but Voodoo doesn't require any additional space (beyond maybe a few meg for files).
Actually, about 200mb, give or take.
Yeah out of the box I believe the internal storage is 1.52GB, with voodoo it drops to 1.31GB. According to appbrain I have 180 apps installed taking up 425mb. So according to the stock task manager I'm using 672mb out of 1.31GB [basically half]. Im not really in danger of running out of space anytime soon
adrynalyne said:
Actually, about 200mb, give or take.
Click to expand...
Click to collapse
Thanks adrynalyne - I stand corrected.
Dang just found out that nfs shift is taking up 100 mb, 100! But is there a way to just relocate this to the memory card?
Sent from my SCH-I500 using XDA App

Best Lagfix for SGS Android 2.3.3/2.3.4

What is Best Lagfix for SGS Android 2.3.3/2.3.4
I dont think you need lagfix anymore on 2.3.3 and 2.3.4. It is already fast, smooth and stable. Though there are some bugs that are not related to lags
Modern lagfixes are mostly just kernels supporting an ext-4 conversion - they all will do more or less the same job. Pick one you like the features of, install and convert.
However as has been mentioned lag has been reduced a lot without it so the improvement is a lot less noticeable than it used to be.
since the GB roms there is no real need for "lagfix"
If you want you can still convert to ext4, just flash any kernel supporting it
Actually there is no noticeable difference between ext4 and rfs fs on gb roms in terms of UI smoothness and speed. And yeah I am using ext4 right now so that is just my personal observation.
jbdroid said:
Actually there is no noticeable difference between ext4 and rfs fs on gb roms in terms of UI smoothness and speed. And yeah I am using ext4 right now so that is just my personal observation.
Click to expand...
Click to collapse
I second this. Lately I've just been using RFS, and I feel it's just as fast as when I ran ext4.
EXT 4 Voodoo lagfix
ext4 rules!
I think ext4 conversion still matters.
I use feedR to follow my google reader feeds. With rfs I can't open posts smoothly and fast as it is with ext4. Actually with rfs it's annoyingly laggy that I can't catch up with all the posts in the feeds.
I'm using feedr to follow 15 websites with rfs, and there is no problem with smoothness or feeds.
Sent from HAL-I9000 using Tapatalk
between rfs n ext4, in GB , yes they r not laggy.
but , how about battery consumption? rfs or ext4 that use alot of battery?
syamsoul said:
between rfs n ext4, in GB , yes they r not laggy.
but , how about battery consumption? rfs or ext4 that use alot of battery?
Click to expand...
Click to collapse
I didnt notice any difference in battery consumption. RFS or EXT4, i usually get 2-3 days of mild usage, 1-2 days of heavy usage, on both.
K0v4L said:
I'm using feedr to follow 15 websites with rfs, and there is no problem with smoothness or feeds.
Sent from HAL-I9000 using Tapatalk
Click to expand...
Click to collapse
Try opening post one after other immediately.
i've been on ext4 ever since eclair, never complained. I keep it like this
Try voodoo lagfix,but on gingerbread i think you don't have to use any lagfix.
Sent from my GT-I9000 using XDA App
Well, my JVH lags as hell...
I don't think you will notice any real difference with different lagfixes. I tried most types and didn't even notice real difference with rfs.
Most devs think ext4 is better so I do apply a lagfix, but not really sure why (better quadrant score? doesn't say anything but still)
I used couple of ROMs from XDA (F1, F1 SGS2, Ficeto, MIUI..) without lagfix... I am using phone a lot (battery lasts max 15-20hrs)... So, There is small diferences between them... I used them without lagfixes... Now, I installed JVP 2.3.4 Stock, Galaxian Kernel, and Converted to EXT4...
Personal opinion:
1. STOCK 2.3.4 JPV - Not bad, Fast, Works fine
2. +Galaxian Kernel - Faster 20-25% at least (using CPU Master because of OC 1.4GHz)
3. EXT4 Convert -> This speeded up my system by 20-30% more...
Explanation.. RFS file system is better with small files (under 5KBytes), and EXT4 is faster and a lot better with larger files... So, when you work with pictures, videos, and applications that uses bigger files then text, EXT4 shows its best side...
Now, my device is focused on Bigger apps, Pictures and HD Videos, so EXT4 is much faster for me...
Use FS that is better for your use...
P.S. EXT4 has faster R/W rate, so when recording, copying and similar, you will see difference...
can't wait until someone gets ubifs file system working, by all accounts should be awesome
Sent using geek power
Am on gb but i use ext4 lagfix because after number of apps increases, it lags!!
I've tried out 2.2 with lagfix and 2.3 without lagfix, but I find that now 2.3 is slower than 2.2 with lagfix. Not sure why this is true, but I noticed that starting apps was slower.

{Q} File system difference?

"All Samsung Roms run on top of BML/RFS, CyanogenMod 7 does NOT.
It runs on MTD/yaffs2 (like Nexus One) which means you'll not able to flash just any kernel or run just any other filesystem you want. Use it as it is if possible, otherwise confirm with the kernel developer that you are trying to install whether it would work with CM. We do not support other kernels and know nothing about their capabilities or compatibility. Only the data partition, which is on movinand, is ext4 like on speedmod or voodoo ("lagfix"). No "lagfix" is necessary because this does not use any Samsung proprietary file systems."
This is a quote from Atinm about the CM7 nightlies.
I'm a bit confused on what it all means.
I know the stock filesystem on our phone is RFS, and then we apply the voodoo lagfix on previous roms to change over to Ext4 to speed things up.
Now what does this mean that the filesystem is now MTD/yaffs2? Is this a better or worse filesystem then EXT4? I'm just curious to know.
yaffs2 is more optimized for flash media than ext4 or rfs/fat. ext4 is more about speed especially on larger drives, but it is media-agnostic. yaffs2 is more focused on data integrity which comes in handy with flash media where you have cycling pages of data that can go bad sometimes. Which isn't to say it's slower either though. Pretty much anything is faster than RFS
So then why was RFS used in the first place? Confusing....
jwleonhart said:
So then why was RFS used in the first place? Confusing....
Click to expand...
Click to collapse
First you have to understand the difference between the partition types.
With flash media, you need something to control wear leveling, whether it be a hardware controller or the filesystem. Bml is essentially a "box" with wear leveling controlled by hardware. This allows you to use common filesystems like RFS (a fat derivative) or voodoo lagfix's ext4. But why Samsung chose this over mtd/yaffs2, no one but the engineers at Samsung knows.
Mtd is the partition type used my the nexus phones, HTC phones, and Google's aosp code. It is essentially a partition of raw flash media, that bypasses any sort of hardware wear leveling. Because of this, you need a filesystems that does this for you, and that's what yaffs2 was designed for. It is not, however, the same as ext4, as ext4 is still meant for conventional mechanical hard drives, al though there are things added in for ssd's but they still require a controller. Yaffs2 is very different different from ext4.
Sent from my SGH-T959 using XDA App
So the question still remains... whats the difference?
Wear leveling would be better do ensure lifr of the internal sd would it not? Does samsung put high quality chip in these phones even?
Sent from my SGH-T959 using XDA Premium App
jwleonhart said:
So the question still remains... whats the difference?
Wear leveling would be better do ensure lifr of the internal sd would it not? Does samsung put high quality chip in these phones even?
Sent from my SGH-T959 using XDA Premium App
Click to expand...
Click to collapse
No, it's not that. It's Tue way current nand flash memory is. That's why there are constantly new controllers coming out for ssds, and constantly new development on filesystems perfectly suited for raw flash memory.
Running a conventional mechanical based filesystems on raw nand would destroy it because it would be constantly writing to the nand memory cell, and you want the exact opposite.
It gets very technical to go any deeper into an explanation, but the reason why CM7 is going with mtd us that their bread and butter (HTC devices, and the nexus series as well as the aosp code) is mtd.
Sent from my SGH-T959 using XDA App
What if it was a filesystem like NTFS? Just for example lets say it could run FAT32.
jwleonhart said:
What if it was a filesystem like NTFS? Just for example lets say it could run FAT32.
Click to expand...
Click to collapse
Ntfs was built for.mechanic hard drives, and would have to rely on a controller to protect the nand flash.
Sent from my SGH-T959 using XDA App
I wouldn't mind a 500gb hard drive in mah phone yooo

Swap File

Most Say that a swap file doesn't make Much of a Difference, From personal experience I would have to disagree
Running free in the terminal on a normal nor swap enabled day I would have 22 - 200 free, this is running no programs, and 3 email accounts
Recently I created and enabled a 512MB swap on my SD, against my better judgment of swapping on solid state..
This has had Drastic improvement in performance let alone free not showing 3916 free in mem and 14424 used on the swap, My shift is alot more fluid and responsive,
Your results may vary but i figured i would share my results
Im running scarystable comp cache 18%, vm heap 48m on CM7 nightly 238
again I don't claim to know anything more then the next guy Just sharing my results
Give me about 15 minutes and you can have these types of stats when you type free in terminal. It does help, regardless of what others have said. And scary kernel will work.
http://db.tt/6NzMLnpl
Sent from my PG06100 using xda premium
I love it, Now my only worry it the wear on my SDcard, but I figure aslong as I delete and create the Swap file once in a while it will ware evenly, I just never expected this kind of gain from it.
I will admit 512 may be overkill but it sounded nice
Thanks for your results!
The problem with swap files/partitions is they're slower then ram so you need to buy the highest class sdcard your device can handle to have anything resembling a positive effect and they'll destroy your sdcard. Plus unless you have a script or app to manage swap its a pain to have to disable and re-enable swap when attaching to your pc through usb.
With our devices having 512 ram and something like V6Supercharger with compcache disabled I've found I always have more then enough free ram for pretty much anything.
Sent from my PG06100 using XDA App
Move thread to here: New thread with instruction and patch
http://forum.xda-developers.com/showthread.php?p=18921074#post18921074
---------- Post added at 01:39 PM ---------- Previous post was at 01:38 PM ----------
Ron Overdrive said:
The problem with swap files/partitions is they're slower then ram so you need to buy the highest class sdcard your device can handle to have anything resembling a positive effect and they'll destroy your sdcard. Plus unless you have a script or app to manage swap its a pain to have to disable and re-enable swap when attaching to your pc through usb.
With our devices having 512 ram and something like V6Supercharger with compcache disabled I've found I always have more then enough free ram for pretty much anything.
Sent from my PG06100 using XDA App
Click to expand...
Click to collapse
No, swap file is slow, swap partition is not
sparksco said:
No, swap file is slow, swap partition is not
Click to expand...
Click to collapse
The data rate of a class 10 sdcard is 10MB/s. DDR-200 which is the oldest and slowest DDR ram is 1600MB/s. Swap is slow regardless whether its a file or a partition.
Sent from my PG06100 using XDA App
Ron Overdrive said:
The data rate of a class 10 sdcard is 10MB/s. DDR-200 which is the oldest and slowest DDR ram is 1600MB/s. Swap is slow regardless whether its a file or a partition.
Sent from my PG06100 using XDA App
Click to expand...
Click to collapse
There is a difference between swap file and swap partition
blahbl4hblah said:
There is a difference between swap file and swap partition
Click to expand...
Click to collapse
Yes, but a swap partition can -not- increase the physical read/write speed of the sdcard which is the point I'm making. Is swap file slower then a partition? Yes I never argued that. My argument is swap in general is slow.
Sent from my PG06100 using XDA App
Ron Overdrive said:
Yes, but a swap partition can -not- increase the physical read/write speed of the sdcard which is the point I'm making. Is swap file slower then a partition? Yes I never argued that. My argument is swap in general is slow.
Sent from my PG06100 using XDA App
Click to expand...
Click to collapse
If your here to use, test swap, provide your results ect fine. If your here to bash swap, then please stfu and go away. We don't need negativity in these forums. do you even know what swappiness is or how to set it?
Swap
Swap is not meant nor does it truly need to be fast, From my understanding it takes whats not being used from ram and "swaps" it to disk to free up ram for programs that require the speed, It is not a replacement for direct ram.
None the less I thank sparksco for his Post and will be trying it out later in much anticipation
JaceAlvejetti said:
Swap is not meant nor does it truly need to be fast, From my understanding it takes whats not being used from ram and "swaps" it to disk to free up ram for programs that require the speed, It is not a replacement for direct ram.
None the less I thank sparksco for his Post and will be trying it out later in much anticipation
Click to expand...
Click to collapse
Your trolling skills amaze me...
Sent from my PG06100 using xda premium
JaceAlvejetti said:
Swap is not meant nor does it truly need to be fast, From my understanding it takes whats not being used from ram and "swaps" it to disk to free up ram for programs that require the speed, It is not a replacement for direct ram.
None the less I thank sparksco for his Post and will be trying it out later in much anticipation
Click to expand...
Click to collapse
Nice
10char
JaceAlvejetti said:
Most Say that a swap file doesn't make Much of a Difference, From personal experience I would have to disagree
Running free in the terminal on a normal nor swap enabled day I would have 22 - 200 free, this is running no programs, and 3 email accounts
Recently I created and enabled a 512MB swap on my SD, against my better judgment of swapping on solid state..
This has had Drastic improvement in performance let alone free not showing 3916 free in mem and 14424 used on the swap, My shift is alot more fluid and responsive,
Your results may vary but i figured i would share my results
Im running scarystable comp cache 18%, vm heap 48m on CM7 nightly 238
again I don't claim to know anything more then the next guy Just sharing my results
Click to expand...
Click to collapse
Swap file is useless and slows your phone down. Just like a pagefile on windows.
"disk" is ALWAYS slower than "RAM"
Post your linpack and antutu scores.

2nd system from webtop partition

I have noticed that the Atrix 2 is capable of running 2nd system from the webtop partition as a form of dual boot, I was wondering if there is any chance of a DEV implementing this into a recovery image, The Atrix 2 uses safestrap that has 2nd system support, was just wondering if it was possible to do this on the Atrix.
Many thanks
kyler084 said:
I have noticed that the Atrix 2 is capable of running 2nd system from the webtop partition as a form of dual boot, I was wondering if there is any chance of a DEV implementing this into a recovery image, The Atrix 2 uses safestrap that has 2nd system support, was just wondering if it was possible to do this on the Atrix.
Many thanks
Click to expand...
Click to collapse
This would be awesome... I don't even have anything in the webtop partition at the moment... and the stupid hardware reserves 200mb of the ram for it anyways... might as well get SOME use out of it!
I'm going to look into it , I have seen it being done on the Razr , I imagine it works in a similar way to the sdcard dualboot just with the mount points pointed at the osh partition.
Sent from my MB860 using xda premium
rossbeck said:
This would be awesome... I don't even have anything in the webtop partition at the moment... and the stupid hardware reserves 200mb of the ram for it anyways... might as well get SOME use out of it!
Click to expand...
Click to collapse
Call me crazy but If the hardware is reserving 200mb of ram for webtop, Wouldnt that be a matter of just editing memory reference.. or would that require the kernel source?
I recall when I had a galaxy 7500GT it had 98 MB of ram and some dev had tailored a kernel to use the 4 mb of video ram for system memory ( at the cost of 3d acceleration) to bump the phone to 102 .. not a huge difference but it helped.
being able to use /osh to boot maybe a super stripped down blur rom for camera/camcorder ect stuff would be nice though.
rossbeck;[URL="tel:31074258" said:
31074258[/URL]]This would be awesome... I don't even have anything in the webtop partition at the moment... and the stupid hardware reserves 200mb of the ram for it anyways... might as well get SOME use out of it!
Click to expand...
Click to collapse
I was under the impression that the 200mb of ram is reserved for the graphics and frame buffers, technically no webtop processes that consume RAM are started unless the device has been docked.
tomh235 said:
I was under the impression that the 200mb of ram is reserved for the graphics and frame buffers, technically no webtop processes that consume RAM are started unless the device has been docked.
Click to expand...
Click to collapse
I think you are right, he might be thinking on the /osh part.
Imo left the system as it is would be better, we all can dual boot from sd and thats good enough. If u want to use use webtop left over space cos of cm7-9-10 then best thing would be to decrease the sise of partition as much as possible. No partition table change no sydtem level mofigications required and u dont have to do any thing especial every time u flash. Correct me if i am wrong
Sent from my MB860 using xda app-developers app
If I didn't need my phone for work, I'd try the sdcard method and symlink to stuff I've flashed into the webtop partition.
Sent from my MB860 using Tapatalk 2

Categories

Resources