Nano 4G

From freemyipod.org
Jump to: navigation, search

Nano 4g frt a.png Nano 4g bck a.png

Components

Label Component Part Markings Notes
2 CPU Samsung S5L8720 339S0049 ARM, K4X56323PI-KGC4, YWE025QH 825, APL0278A00, N1B2HOP 0831 ARM1176JZF-S processor. It is definitely worth knowing that this is the exact same processor used in the iTouch 2G. This could mean that some of the same exploits for that could possibly be used. Here is a very interesting page about the S5L8720 processor.
SDRAM 32MB, probably MDDR. Integrated into the processor, similar to the iPod Touch and iPhone lines.
4 Accelerometer LIS302DL 33DL, 2827 The newer Touch's, iPhone's, and even the iPad have similar accelerometers, and I've discovered a pattern in the chip names.
6 NAND Flash Varies TH58NVG6D1DLA87, U20516, JAPAN, 0826MAE
5 Audio codec CS42L58 338S055C, 189N0824, SGP I determined this because the Nano 5G has a similar chip, which we are sure of the identity. One person lifted this chip and found that the pins connect to the LCD connector. Not much info was given, and it could just be a common ground, but the identity of this chip is still up in the air.
1 Power manager Dialog D1759 338S0687-AC, 08288HBB
3

Bootrom

See S5L8720 Bootrom. Different from the S5L8720 bootrom used in the iPod Touch 2G (which is iBoot-based, a.k.a. SecureROM).

Memory Map

See [1] and [2].

In addition to the above, a few extra memory regions have been found while reverse engineering the S5L8720 Bootrom:

Name Address Notes
Mystery DMA 0x3880_0000 A PL080-like DMA engine, but with slightly different MMIO register structure. Used by the bootrom to copy the DFU payload from 0x2200_0600 to 0x2200_0000 after decryption and verification. Or maybe that's actually doing the decryption? To be investigated.
Mystery Interrupt Thing 0x39a0_0000 Not the VICs (0x38e0_0000, 0x38e0_1000), not the EdgeIC (0x38e0_2000). Seems to hold 7 different 32-bit registers for interrupt status at 0xa0, and 7 different 32-bit registers for interrupt mask at 0xc0. The 7 different registers correspond to 7 'modes' of ISRs set up in the bootrom. Not much is known about what it does, and what these 'modes' are. To be investigated.

Reverse Engineering Results

Timers: These clockgates have been found to be related to timers: 37, 55, 56, 57, 58, 59, 60, 69, 70, 128, 129, 130, 131, 132, 133, 134, 150 and 151.

Status registers

We dumped all c0 coprocessor registers:

c0,c0

Value: 0x410FB764

Interpretation: ARM1176 rev. 4

c0,c1

Value: 0x1D152152

Interpretation: DCache/ICache 16KB each, 4 way associative, 32 bytes line size

c0,c2

Value: 0x00000000

Interpretation: No TCM

c0,c3

Value: 0x00000800

Interpretation: Unified TLB, 8 lockable entries

c1,c0

Value: 0x00000111

Interpretation: ARM/Thumb1/Jazelle support, no Thumb2 support

c1,c1

Value: 0x00000011

Interpretation: Trustzone v1

c1,c2

Value: 0x00000033

Interpretation: Supports debug model v6.1, both applications processor and secure

c1,c3

Value: 0x00000000

Interpretation: No auxiliary features

c1,c4

Value: 0x01130003

Interpretation: FCSE, Auxiliary Control register, ARMv6 TCM/DMA, no DMA cache coherency, no multicore cache coherency, VMSA v7

c1,c5

Value: 0x10030302

Interpretation: Branch target buffer, Harvard architecture, various cache operations supported (see TRM)

c1,c6

Value: 0x01222100

Interpretation: WFI, Data synchronization barrier, Prefetch flush, Data memory barrier, various TLB/cache operations supported (see TRM), no prefetch cache range operation

c1,c7

Value: 0x00000000

Interpretation: No hierarchical cache maintenance support

c2,c0

Value: 0x00140011

Interpretation: Supports BKPT, CDP, CDP2, LDC, LDC2, MCD, MCD2, MRC, MRC2, STC, STC2, MCRR, MCRR2, MRRC, MRRC2, CLZ, SWP and SWPB, doesn't support division, combined compare and branch or bitfield instructions

c2,c1

Value: 0x12002111

Interpretation: Supports BXJ, BX, BLX, PC loads have BX behavior, supports SXTB, SXTAB, SXTB16, SXTAB16, SXTH, SXTAH, UXTB, UXTAB, UXTB16, UXTAB16, UXTH, UXTAH, SRS, RFE, CPS, LDM(2), LDM(3), STM(2) and SETEND

c2,c2

Value: 0x11231121

Interpretation: Supports REV, REV16, REVSH, MRS, MSR, UMULL, UMLAL, UMAAL, SMULL, SMLAL, SMLABB, SMLABT, SMLALBB, SMLALBT, SMLALTB, SMLALTT, SMLATB, SMLATT, SMLAWB, SMLAWT, SMULBB, SMULBT, SMULTB, SMULTT, SMULWB, SMULWT, SMLAD, SMLADX, SMLALD, SMLALDX, SMLSD, SMLSDX, SMLSLD, SMLSLDX, SMMLA, SMMLAR, SMMLS, SMMLSR, SMMUL, SMMULR, SMUAD, SMUADX, SMUSD, SMUSDX, MLA, restartable LDM/STM, PLD, LDRD, STRD and Q bit in PSRs

c2,c3

Value: 0x01102131

Interpretation: Supports true NOP, Thumb MOV(3)/CPU, LDREX, STREX, LDREXB, LDREXH, LDREXD, STREXB, STREXH, STREXD, CLREX, SVC, PKHBT, PKHTB, QADD16, QADD8, QADDSUBX, QSUB16, QSUB8, QSUBADDX, SADD16, SADD8, SADDSUBX, SEL, SHADD16, SHADD8, SHADDSUBX, SHSUB16, SHSUB8, SHSUBADDX, SSAT, SSAT16, SSUB16, SSUB8, SSUBADDX, SXTAB16, SXTB16, UADD16, UADD8, UADDSUBX, UHADD16, UHADD8, UHADDSUBX, UHSUB16, UHSUB8, UHSUBADDX, UQADD16, UQADD8, UQADDSUBX, UQSUB16, UQSUB8, UQSUBADDX, USAD8, USADA8, USAT, USAT16, USUB16, USUB8, USUBADDX, UXTAB16, UXTB16, QADD, QDADD, QDSUB, QSUB, and the Q and GE[3:0] bits in the PSRs. Does nut support branch table and Thumb2 instructions.

c2,c4

Value: 0x00001141

Interpretation: Supports SMC, writeback instructions, shift of loads and stores by 0-3 bits to the left, constant shift options, register controlled shift options, LDRBT, LDRT, STRBT and STRT. No barrier instructions support.

c2,c5

Value: 0x00000000

Interpretation: No additional implementation defined instruction set extensions

Helpful pages

Teardowns:

Other: