Make the client submit unbatched transactions when the hardware wallet is used.

Author: murisi

crates/apps_lib/src/client/tx.rs

@@ -258,7 +258,7 @@ async fn batch_opt_reveal_pk_and_submit<N: Namada>(
     namada: &N,
     args: &args::Tx,
     owners: &[&Address],
-    tx_data: (Tx, SigningTxData),
+    mut tx_data: (Tx, SigningTxData),
 ) -> Result<ProcessTxResponse, error::Error>
 where
     <N::Client as namada_sdk::io::Client>::Error: std::fmt::Display,
@@ -272,15 +272,33 @@ where
             batched_tx_data.push(reveal_pk_tx_data);
         }
     }
-    batched_tx_data.push(tx_data);
 
-    let (mut batched_tx, batched_signing_data) =
-        namada_sdk::tx::build_batch(batched_tx_data)?;
-    for sig_data in batched_signing_data {
-        sign(namada, &mut batched_tx, args, sig_data).await?;
+    // Since hardware wallets do not yet support batched transactions, use a
+    // different logic in order to achieve compatibility
+    if args.use_device {
+        // Sign each transaction separately
+        for (tx, sig_data) in &mut batched_tx_data {
+            sign(namada, tx, args, sig_data.clone()).await?;
+        }
+        sign(namada, &mut tx_data.0, args, tx_data.1).await?;
+        // Then submit each transaction separately
+        for (tx, _sig_data) in batched_tx_data {
+            namada.submit(tx, args).await?;
+        }
+        // The result of submitting this function's argument is what is returned
+        namada.submit(tx_data.0, args).await
+    } else {
+        // Otherwise complete the batch with this function's argument
+        batched_tx_data.push(tx_data);
+        let (mut batched_tx, batched_signing_data) =
+            namada_sdk::tx::build_batch(batched_tx_data)?;
+        // Sign the batch with the union of the signers required for each part
+        for sig_data in batched_signing_data {
+            sign(namada, &mut batched_tx, args, sig_data).await?;
+        }
+        // Then finally submit everything in one go
+        namada.submit(batched_tx, args).await
     }
-
-    namada.submit(batched_tx, args).await
 }
 
 pub async fn submit_bridge_pool_tx<N: Namada>(
@@ -1202,7 +1220,7 @@ pub async fn submit_shielded_transfer(
         MAX_HW_OUTPUT,
         &args.tx,
     )
-    .await?;
+        .await?;
     let (mut tx, signing_data) =
         args.clone().build(namada, &mut bparams).await?;
     masp_sign(&mut tx, &args.tx, &signing_data, shielded_hw_keys).await?;
@@ -1378,7 +1396,7 @@ where
         MAX_HW_OUTPUT,
         &args.tx,
     )
-    .await?;
+        .await?;
     let (mut tx, signing_data, _) = args.build(namada, &mut bparams).await?;
     masp_sign(&mut tx, &args.tx, &signing_data, shielded_hw_keys).await?;
 
@@ -1396,7 +1414,7 @@ where
             &[&args.source.effective_address()],
             (tx, signing_data),
         )
-        .await?;
+            .await?;
 
         if let Some(masp_section) = opt_masp_section {
             pre_cache_masp_data_on_tx_result(namada, &res, &masp_section).await;

crates/tests/src/e2e/ibc_tests.rs

@@ -1403,7 +1403,7 @@ fn transfer_on_chain(
         &rpc,
     ]);
     tx_args.extend_from_slice(extra_args);
-    let mut client = run!(test, Bin::Client, tx_args, Some(120))?;
+    let mut client = run!(test, Bin::Client, tx_args, Some(240))?;
     client.exp_string(TX_APPLIED_SUCCESS)?;
     client.assert_success();