User Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:30.0) Gecko/20100101 Firefox/30.0 (Beta/Release) Build ID: 20140605174243

this is the related HCI Event - EVT-TRANSACTION - logs

Accroding to ETSI TS 102 622 specification, the correct data structure of the EVT-TRANSACTION is made of following two parts of TLV(Tag-Length-Value) objects. TAG(1 byte) Len(1 byte) Description ------------------------------------------------- 0x81 5-16 AID 0x82 0-255 Parameters please see the document on the page 55, section 11.2.2.4 http://www.etsi.org/deliver/etsi_ts/102600_102699/102622/12.00.00_60/ts_102622v120000p.pdf On the Flame device, the NFCd received following EVT-TRANSACTION data from the underlying Vendor NFC library(BRCM-NCI and NXP-HAL), which is incorrect! The wrong event data on Flame with some duplicated info: 0x81 <aid-len> <aid> 0x82 0x81 <aid-len> <aid> 0x82 <params-len><params> The expected event data on Flame: 0x81 <aid-len> <aid> 0x82 <params-len><params>

Please refer to the detailed log info about incorrect hci-event "EVT-TRANSACTION" collected on Flame. The log file: https://bugzilla.mozilla.org/attachment.cgi?id=8452220

Hi Wesley, From the log, libnfc-nci pass incorrect data format of HCI_EVT_TRANSACTION to nfcd. We will need vendor's help to check it.

calling out to Francis for raising this issue to vendor.

hi Yanan, youlong, could you please check comment2: On the Flame device, the NFCd received following EVT-TRANSACTION data from the underlying Vendor NFC library(BRCM-NCI and NXP-HAL), which is incorrect! please kindly check with NXP and clarify this issue. thank you very much

We've done some test of HCI EVT_TRANSACTION with our test SIM cards and TMPL commercial SIM. We were using todays build and Dimi's patches from Bug 979891 with additional logging. All the tests went well. With the help of TMPL equipment we have been also sniffing SWP communication, EVT_TRANSACTION content logged in nfcd was matching the data sent via SWP. In the attachment you can see adb logcat for payment in a vending machine, event transaction log starts around line 3505. In all tests we haven't seen any duplicated data. Maybe there is something wrong with the applet on Dimi's SIM card?

Thanks for the Krzysztof's hci-payment.log done with commercial SIM cards and the SIM applet that issues can also issue the EVT-TRANSACTION data, I am aware now that I made the mistake in my test SIM applet, which sends out the incorrect EVT-TRANSACTION data to NFC Controller. In a nutshell, Vendor NFC lib doesn't make any mistake on this point. Please withdraw the request from t2mobile.

(In reply to Francis Lee [:frlee] from comment #6) > hi Yanan, youlong, > > could you please check comment2: > > On the Flame device, the NFCd received following EVT-TRANSACTION data from > the underlying Vendor NFC library(BRCM-NCI and NXP-HAL), which is incorrect! > > please kindly check with NXP and clarify this issue. > > thank you very much Hi Francis, I have identified the reason for the previous reported HCI-event problem. It is not the fault on the vendor side, instead the problem comes from the SIM application side. Now, it has been fixed. Could you please withdraw the request from T2mobile? thanks anyway for your support!

hi Ming, thanks a lot for your quick update. hi Yanan, Youlong, as comment9, please ignore this issue. and thank you for your help.