data.memory.bank.schedule ≡

template <auto Banks, typename Address, auto Addresses>
inline 
count_t<Addresses>[Banks]
read_requests_per_bank(
    Address[Addresses] addresses,
    (Address) -> index_t<Banks> address_to_bank_index
    ) §source

Count the number of unique read requests for each bank. More than 1 request per bank indicates conflict.

Parameters

```
auto Banks
```
```
typename Address
```
```
auto Addresses
```
Number of addresses processed per call.

Arguments

```
Address[Addresses] addresses
```
Input addresses, duplicates allowed.

(Address) -> index_t<Banks> address_to_bank_index

Map Address to bank index.

template <auto Banks, typename Address, auto Addresses>
inline 
count_t<Addresses>[Banks]
write_requests_per_bank(
    optional<Address>[Addresses] addresses,
    (Address) -> index_t<Banks> address_to_bank_index
    ) §source

Count the number of write requests for each bank. More than 1 request per bank indicates conflict.

Parameters

```
auto Banks
```
```
typename Address
```
```
auto Addresses
```
Number of addresses processed per call.

Arguments

```
optional<Address>[Addresses] addresses
```
Input addresses, valid duplicates not allowed.

(Address) -> index_t<Banks> address_to_bank_index

Map Address to bank index.

template <auto Banks, typename Address, auto Addresses>
inline 
pair<optional<index_t<Addresses>>[Banks], optional<index_t<Banks>>[Addresses]>
schedule_read_requests(
    Address[Addresses] addresses,
    (Address) -> index_t<Banks> address_to_bank_index,
    index_t<Addresses> iteration
    ) §source

For each iteration, calculate the pair:

The optional address indices for each bank.
The optional bank index for each address.

Parameters

```
auto Banks
```
```
typename Address
```
```
auto Addresses
```
Number of addresses processed per call.

Arguments

```
Address[Addresses] addresses
```
Input addresses, duplicates allowed.

(Address) -> index_t<Banks> address_to_bank_index

Map Address to bank index.

```
index_t<Addresses> iteration
```
Index of bank requests to retrieve from schedule.

template <auto Banks, typename Address, auto Addresses>
inline 
optional<index_t<Addresses>>[Banks]
schedule_write_requests(
    optional<Address>[Addresses] addresses,
    (Address) -> index_t<Banks> address_to_bank_index,
    index_t<Addresses> iteration
    ) §source

Calculate the optional address indices for each bank at specific iteration.

Parameters

```
auto Banks
```
```
typename Address
```
```
auto Addresses
```
Number of addresses processed per call.

Arguments

```
optional<Address>[Addresses] addresses
```
Input addresses, valid duplicates not allowed.

(Address) -> index_t<Banks> address_to_bank_index

Map Address to bank index.

```
index_t<Addresses> iteration
```
Index of bank requests to retrieve from schedule.

Index

data.memory.bank.schedule ≡