[PATCH 2/4] cmd: mvebu/bubt: Correctly propagate failure during tftp transport

[Pali Rohár]

net_loop() returns signed int type and negative value represents error.
tftp_read_file() returns unsigned size_t type and zero value represents
error. Casting signed negative value to unsigned size_t type cause losing
information about error and bubt thinks that no error happened, and
continue erasing SPI-NOR which cause malfunction device.

Fix this issue by correctly propagating failure during tftp transport.

With this change when there is no eth link, bubt does not erase SPI-NOR
anymore.

  => bubt
  Burning U-Boot image "flash-image.bin" from "tftp" to "spi"
  ethernet at 30000 Waiting for PHY auto negotiation to complete......... TIMEOUT !
  ethernet at 30000: No link.
  Error: Failed to read file flash-image.bin from tftp

Signed-off-by: Pali Rohár <pali at kernel.org>
---
 cmd/mvebu/bubt.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/cmd/mvebu/bubt.c b/cmd/mvebu/bubt.c
index a05e22a5479c..2924b1539f32 100644
--- a/cmd/mvebu/bubt.c
+++ b/cmd/mvebu/bubt.c
@@ -455,11 +455,14 @@ static int is_usb_active(void)
 #ifdef CONFIG_CMD_NET
 static size_t tftp_read_file(const char *file_name)
 {
+	int ret;
+
 	/*
 	 * update global variable image_load_addr before tftp file from network
 	 */
 	image_load_addr = get_load_addr();
-	return net_loop(TFTPGET);
+	ret = net_loop(TFTPGET);
+	return ret > 0 ? ret : 0;
 }
 
 static int is_tftp_active(void)
-- 
2.20.1