I'll tell you what I learned about
Sudden death syndrome /
Sudden Death Syndrome /
SDS On samsung gt-i9300 Galaxy S3.
First of all, I note that Samsung did not make any official statements about SDS. Somehow the questions-response of Samsung employee in the media, I do not take into account. However, on our forum there are messages with the problems of reading and writing to the internal flash memory.
16Gb When firmware (for example, ODIN PC cannot receive information about posting sections - freezes on Get Pit for Mapping). Extreme case - the device is not loaded.
It is known that samsung is periodically
Publishes the source codeFirmware, as requires a licensed Google policy for Android. So, in the publishing code
Update7.A certain strange fix was found (patch / patch / program code correcting the problem).
What makes this fix? With each initialization of the inner flash drive, it resets part of the flash memory controller code (VTU00M - controller, EMMC or Embedded Multimediacard - flash memory, and all together, figuratively speaking - disk). It is very similar to solving the problem of our SDS in the form of a preventive measure, although Samsung officially did not declare SDS fixes, as did not declare the fact of the presence of SDS itself.
Judging by the fix code, there is a cumulative error when working with a flash memory controller, which Fix constantly resets when the flash memory is initialized. Fix does not write a firmware into the flash memory controller, the Fix does not put the filter on accessing flash memory - Fix constantly reset the controller data. Thus, this procedure is not very correctly called Fix as such, but by solving the problem (I hope is successful). However, everyone got used to the "SDS Fix", so let Fix remain.
Running forward, voice
Opinion from the XDA Forum: During operation, a flush memory blocks can occur, and the SDS fix prevents this damage.
Fix is ​​located in Kernel (kernel), therefore must be present both in the Boot of the main firmware and in the boot recovery. The list where the firmware containing this fixes and parts of the firmware are constantly updated here:
[Important] Sudden Death Fix - Are You Covered?.
My studies source code Update7
1. When initializing the internal card, the name VTU00M and the audit of the F1 chip are checked.
2. Fix 4 control commands to verify the presence of a return error or for what purpose.
3. From the memory read the date.
4. The date from 2012.04.13 is required (with what comparison is: if (date! =0x 20120413))
5. If the chip, revision and date coincide, then MMC_START_MOVI_OPERATION is executed (this subroutine is indicated as exported)
6. In MMC_START_MOVI_OPERATION, I could only understand the fact that each time the chip memory is clogged with zeros.
It's all.
Regarding the presence of SDS fixes from Samsung
Update7.In the official firmware 4.1.2 for Russia i9300xxe
LL4. ChangeList: 566833, Date: 12/10/2012.
1. Let me remind you, Fix appeared in the open UPDATE7 code. Come
diffpresented
AndreiLux(MMC: Soft-Patch Movinand VTU00M (16GB) EMMC Failure).
2. Fixes have specific text strings "CMD13 (VC) Failed" and "VTU00M", which we can independently find in binary files without complicated passages with reengineering.
case history
At first, one craftsman took the text strings from MMC_OPS.C and MMC.C from Update7 "Printk (Kern_err" CMD13 (VC) Failed \ N ")" and "if (! StrnCMP (HOST->Card->CID.prod_name, "VTU00M", 6) ", and found them in LLA. Then research on XDA has shown that the traces of the Update7 code are in all firmware 4.1.2
3. To independently find a boot.img file. I checked on the file taken from code_i9300xxell4_566833_rev00_user_low_ship.tar.md5
Ambulance. To search in custom, the method below may not work (random coincidences of the scholar codes). This method for Waste is accurately working.
4. The hex editor will be needed. I took
Notepad ++and
Hex Editor Plugin.to him.
5. Open boot.img in the Hex editor.
6. Find the first match of 1F 8B (start GZIP)
7. Delete everything from addresses 0 to 1F8B (1F8B will be the beginning of our new file)
8. Find the string 00 1F 8B (in some customs it is not the first, but the second coincidence)
9. Remove everything after 00 from 1F8B to the end of the file (our 00 will be the end of the new file)
10. Save the GZ extension file. I saved like boot.img.gz
11. Either
7zipOr using Total Commander (I checked on v7.56a) Remove the file (my WinRAR did not cope with the archive)
Open the extracted file and look for the VTU00M and CMD13 (VC) Failed line there. If found, then you have Fix.
I found Fixes Fixes in i9300xxell4 SER:
VTU00m.<4>% S: MOVI OPERATION IS FAILED
<3>CMD13 (VC) Failed
And if so, then in the Russian official firmware i9300xxell4 Ser Fix is ​​there! So I confirm research
XDA! Just in case I checked i9300xxe
MB5. who already belongs to
Update9. - Fix SDS will not go anywhere.
By the way, you can also check the image of the recovery - recovery.img
No longer doubt that Fix SDS is available in all the firmware 4.1.2 without exception. This means that i9300 with firmware 4.1.2 and above
not Must be subject to SDS. SDS Fixes No. 4.0.4 and 4.1.1 firmware.
We are available for analysis two versions of the Fix Code - from
Update7.and
Update9.. It is written that Update9 corresponds to the firmware XXEMB5. It is also written there that update7 corresponds to xxella, but if the SDS fix was found in XXELL4, then it is not necessary to unconditionally verify administrators OpenSource.samsung.com.
In the message
Ultimate Galaxy S3 Unusual Freezing ThreadIt is assumed that without the SDS fixes, the blocks of flash memory deteriorate, and the SDS fixes do not deteriorate, but the phone can be filled with the already spoiled block. It is also indicated that from the XXEmb5 version, the problem of fading was solved with the new version of the Fix SDS ... Well, maybe, I do not know. In the update9 code, I did not understand this, but I do not pretend to the programmer's title. I do not draw conclusions.
In the same
the messageA method for fixing bad blocks is proposed. You need to install the firmware having the Fix SDS, and several times completely score flash memory by any data in order to mark the bad blocks during this operation, and not at the moment when you urgently need a smartphone, because this operation can hire 5 -20 minutes on the block. On there is attached.
List of firmware 4.1.2 on April 24, 2013
I9300XXELKC, I9300XXELL1, I9300XXELL4, I9300XXELL5, I9300XXELL6, I9300XXELLA, I9300XXELLB, I9300XXELLC, I9300ZSEMA1, I9300XXEMA2, I9300XXEMB2, I9300XXEMB2, I9300XXEMB5, I9300XXEMR2, I9300ZSEMC1, I9300ZSEMC2, I9300ZSEMC3. The list does not claim the title of complete, but the firmware mentioned in this theme is there.
At the moment I know:
1.
Mobile Odin v3.65.When the firmware uses its temporary recovery - Flashkernels having Fix SDS.
2.
TWRP 2.4.x.x Touch RecoveryIt has a fix SDS.
***
Thinking out loud.
If SDS scares you, then hurt any official firmware version 4.1.2, and it is better to firmware not lower than XXEMB5. It will not be worse! How to understand which firmware is newer, detailed in detail in
by thismessage.
You can rout the device - the fact of rutting on the SDS does not affect. The CPU overclocking can affect, change sections, and a lot of things can be done under root, but this is another topic.
Bootloader ... if you come from logic, then in new bootloaders, SDS protection should be. At the same time, no one has yet been able to prove that in the bootloader there is a Fix SDS, in contrast to the obvious evidence of the availability of the SDS fixture in the kernel. Establish you, put a new bootloader or leave the old one. I left the old one.
Statistics on this forum shows that the overwhelming number of those who went to the service center had on board the firmware smartphone 4.0.4 and 4.1.1. The SC recorded problems with flash memory and changed the motherboard. Repair cost from 8 to 12 thousand rubles. So is it worth putting 4.0.4 and 4.1.1 if I have 4.1.2? ..
Post has been editedVetal`M - 08.05.17, 14:41Reason for editing: In the header