mirror of
https://github.com/RfidResearchGroup/proxmark3.git
synced 2026-10-23 08:43:07 +00:00
Refactor comments in iso14a_rebase_transfer_time
Updated comments for clarity regarding NextTransferTime handling. Signed-off-by: Antiklesys <syselkitna@gmail.com>
This commit is contained in:
1 parent
9479340e19
commit
aadefed412
1 file changed
+2
-4
+2
-4
@@ -315,10 +315,8 @@ uint32_t iso14a_get_timeout(void) {
|
||||
}
|
||||
|
||||
void iso14a_rebase_transfer_time(void) {
|
||||
// NextTransferTime is expressed in absolute SSP-clock counts. A caller
|
||||
// that restarts StartCountSspClk() resets that counter to zero; retaining
|
||||
// the old value would make ReaderTransmit() wait for stale time from the
|
||||
// preceding APDU. Do not reset PCB or ATS state here.
|
||||
//Restarting StartCountSspClk() resets NextTransferTime clock-counts to zero; retaining old value makes ReaderTransmit() wait stale time from previous APDU.
|
||||
//Do not reset PCB or ATS state here.
|
||||
NextTransferTime = 2 * DELAY_ARM2AIR_AS_READER;
|
||||
}
|
||||
|
||||
|
||||
Reference in new issue
Block a user