Hello Peter Hurley,
The patch 33d9b8b23a73: "serial: 8250: Extract IIR logic steering
from rx dma" from Apr 9, 2016, leads to the following static checker
warning:
drivers/tty/serial/8250/8250_omap.c:1093 handle_rx_dma()
warn: signedness bug returning '(-22)'
drivers/tty/serial/8250/8250_omap.c
1084 static bool handle_rx_dma(struct uart_8250_port *up, unsigned int iir)
^^^^
This returns bool
1085 {
1086 switch (iir & 0x3f) {
1087 case UART_IIR_RLSI:
1088 case UART_IIR_RX_TIMEOUT:
1089 case UART_IIR_RDI:
1090 omap_8250_rx_dma_flush(up);
1091 return true;
1092 }
1093 return omap_8250_rx_dma(up);
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This returns error codes
1094 }
1095
1096 static unsigned char omap_8250_handle_rx_dma(struct uart_8250_port *up,
1097 u8 iir, unsigned char status)
1098 {
1099 if ((status & (UART_LSR_DR | UART_LSR_BI)) &&
1100 (iir & UART_IIR_RDI)) {
1101 if (handle_rx_dma(up, iir)) {
^^^^^^^^^^^^^^^^^^^^^^
So I guess this is reversed for the omap_8250_rx_dma() case.
1102 status = serial8250_rx_chars(up, status);
1103 omap_8250_rx_dma(up);
1104 }
1105 }
1106
1107 return status;
1108 }
regards,
dan carpenter
