| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Tue, 6 Feb 2007 18:39:57 +0800
From: "Aubrey Li" <aubreylee@...il.com>
To: "Wu, Bryan" <bryan.wu@...log.com>, alan@...rguk.ukuu.org.uk,
aubreylee@...il.com, akpm@...l.org, rdunlap@...otime.net,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 2/3, try #2] Blackfin: serial driver for Blackfin architecture against Linux kernel 2.6.20
On 2/6/07, Russell King <rmk+lkml@....linux.org.uk> wrote:
> I think you have a bit of work to do on this driver; it seems to be
> missing all termios handling apart from setting the baud rate.
Yes, only baud rate setting is enabled. We'll implement others in future.
> Moreover, it's re-using the 8250 driver port range despite claiming
> to have a proper allocation.
Thanks to point this out, although 8250 driver and bfin serial driver
will never be enabled at the same time.
Is there a port map to show which ports are used? Or how do I know
which port is available?
>
> (I don't maintain serial anymore but I do reserve the right to
> occasionally review serial code.)
>
> On Tue, Feb 06, 2007 at 11:20:11AM +0800, Wu, Bryan wrote:
> > +/* We've been assigned a range on the "Low-density serial ports" major */
> > +#define SERIAL_BFIN_MAJOR TTY_MAJOR
> > +#define MINOR_START 64
>
> If you've been assigned a range in the low density serial ports major,
> why aren't you using it instead of taking the 8250 drivers range (thereby
> preventing PCMCIA cards from ever being used) ?
>
> > +#if defined(CONFIG_BAUD_9600)
> > +#define CONSOLE_BAUD_RATE 9600
> > +#elif defined(CONFIG_BAUD_19200)
> > +#define CONSOLE_BAUD_RATE 19200
> > +#elif defined(CONFIG_BAUD_38400)
> > +#define CONSOLE_BAUD_RATE 38400
> > +#elif defined(CONFIG_BAUD_57600)
> > +#define CONSOLE_BAUD_RATE 57600
> > +#elif defined(CONFIG_BAUD_115200)
> > +#define CONSOLE_BAUD_RATE 115200
> > +#endif
>
> What's wrong with passing console=ttyblackfin0,115200 to the kernel or
> via some default command line?
We are not used to pass parameter like that. Put it in Kconfig option
is more friendly to our customer.
> > +/*
> > + * interrupts are disabled on entry
> > + */
> > +static void bfin_serial_stop_tx(struct uart_port *port)
> > +{
> > + struct bfin_serial_port *uart = (struct bfin_serial_port *)port;
> > +
> > +#ifdef CONFIG_SERIAL_BFIN_DMA
> > + disable_dma(uart->tx_dma_channel);
>
> It's far better to stop transmission as soon as possible to prevent
> overruns at the remote end. I doubt disabling DMA achieves that.
UART peripheral on blackfin has two work mode, one is PIO, the other is DMA.
The driver enables two mode. If overruns at the remote end is critical
for someone, PIO is a better choice.
>
> > +#else
> > + unsigned short ier;
> > +
> > + ier = UART_GET_IER(uart);
> > + ier &= ~ETBEI;
> > + UART_PUT_IER(uart, ier);
> > +#endif
> > +}
> ...
> > +static void bfin_serial_rx_chars(struct bfin_serial_port *uart)
> > +{
> > + struct tty_struct *tty = uart->port.info?uart->port.info->tty:0;
> > + unsigned int status, ch, flg;
> > +#if defined(CONFIG_BF531) || defined(CONFIG_BF532) || defined(CONFIG_BF533)
> > + static int in_break = 0;
> > +#endif
> > +
> > + status = UART_GET_LSR(uart);
> > + ch = UART_GET_CHAR(uart);
> > + uart->port.icount.rx++;
> > +
> > +#if defined(CONFIG_BF531) || defined(CONFIG_BF532) || defined(CONFIG_BF533)
> > + if (in_break) {
> > + if (ch != 0) {
> > + in_break = 0;
> > + ch = UART_GET_CHAR(uart);
> > + }
> > + return;
> > + }
> > +#endif
> > +
> > + if (status & BI) {
> > +#if defined(CONFIG_BF531) || defined(CONFIG_BF532) || defined(CONFIG_BF533)
> > + in_break = 1;
> > +#endif
> > + uart->port.icount.brk++;
> > + if (uart_handle_break(&uart->port))
> > + goto ignore_char;
> > + flg = TTY_BREAK;
> > + } else if (status & PE) {
> > + flg = TTY_PARITY;
> > + uart->port.icount.parity++;
> > + } else if (status & OE) {
> > + flg = TTY_OVERRUN;
> > + uart->port.icount.overrun++;
> > + } else if (status & FE) {
> > + flg = TTY_FRAME;
> > + uart->port.icount.frame++;
> > + } else
> > + flg = TTY_NORMAL;
>
> Why do many serial drivers when they're first written utterly ignore
> termios modes in their reception path by not using port.read_status_mask?
> It's broken as stands.
>
> > +
> > + if (uart_handle_sysrq_char(&uart->port, ch))
> > + goto ignore_char;
> > + if (tty)
> > + uart_insert_char(&uart->port, status, 2, ch, flg);
> > +
> > +ignore_char:
> > + if (tty)
> > + tty_flip_buffer_push(tty);
> > +}
> ...
> > +static irqreturn_t bfin_serial_int(int irq, void *dev_id)
> > +{
> > + struct bfin_serial_port *uart = dev_id;
> > + unsigned short status;
> > +
> > + spin_lock(&uart->port.lock);
> > + status = UART_GET_IIR(uart);
> > + do {
> > + if ((status & IIR_STATUS) == IIR_TX_READY)
> > + bfin_serial_tx_chars(uart);
> > + if ((status & IIR_STATUS) == IIR_RX_READY)
> > + bfin_serial_rx_chars(uart);
> > + status = UART_GET_IIR(uart);
> > + } while (status &(IIR_TX_READY | IIR_RX_READY));
>
> Space between '&' and '(' please.
>
> > + spin_unlock(&uart->port.lock);
> > + return IRQ_HANDLED;
> > +}
> ...
> > +static void bfin_serial_dma_rx_chars(struct bfin_serial_port * uart)
> > +{
> > + struct tty_struct *tty = uart->port.info->tty;
> > + int i, flg, status;
> > +
> > + status = UART_GET_LSR(uart);
> > + uart->port.icount.rx += CIRC_CNT(uart->rx_dma_buf.head, uart->rx_dma_buf.tail, UART_XMIT_SIZE);;
> > +
> > + if (status & BI) {
> > + uart->port.icount.brk++;
> > + if (uart_handle_break(&uart->port))
> > + goto dma_ignore_char;
> > + flg = TTY_BREAK;
> > + } else if (status & PE) {
> > + flg = TTY_PARITY;
> > + uart->port.icount.parity++;
> > + } else if (status & OE) {
> > + flg = TTY_OVERRUN;
> > + uart->port.icount.overrun++;
> > + } else if (status & FE) {
> > + flg = TTY_FRAME;
> > + uart->port.icount.frame++;
> > + } else
> > + flg = TTY_NORMAL;
>
> No termios handling.
>
> > +
> > + for (i = uart->rx_dma_buf.head; i < uart->rx_dma_buf.tail; i++) {
> > + if (uart_handle_sysrq_char(&uart->port, uart->rx_dma_buf.buf[i]))
> > + goto dma_ignore_char;
> > + uart_insert_char(&uart->port, status, 2, uart->rx_dma_buf.buf[i], flg);
> > + }
> > +dma_ignore_char:
> > + tty_flip_buffer_push(tty);
> > +}
> ...
> > +/*
> > + * Handle any change of modem status signal since we were last called.
> > + */
> > +static void bfin_serial_mctrl_check(struct bfin_serial_port *uart)
> > +{
> > +#ifdef CONFIG_SERIAL_BFIN_CTSRTS
> > + unsigned int status;
> > +# ifdef CONFIG_SERIAL_BFIN_DMA
> > + struct uart_info *info = uart->port.info;
> > + struct tty_struct *tty = info->tty;
> > +
> > + status = bfin_serial_get_mctrl(&uart->port);
> > + if (!(status & TIOCM_CTS)) {
> > + tty->hw_stopped = 1;
> > + } else {
> > + tty->hw_stopped = 0;
> > + }
>
> A change of CTS doesn't always mean transmission has to stop.
> uart_handle_cts_change() knows when this has to happen. Use it.
If I recall correctly, uart_handle_cts_change can't handle DMA case on blackfin.
>
> > +# else
> > + status = bfin_serial_get_mctrl(&uart->port);
> > + uart_handle_cts_change(&uart->port, status & TIOCM_CTS);
> > + if (!(status & TIOCM_CTS))
> > + schedule_work(&uart->cts_workqueue);
> > +# endif
> > +#endif
> > +}
> ...
> > +static void
> > +bfin_serial_set_termios(struct uart_port *port, struct ktermios *termios,
> > + struct ktermios *old)
> > +{
> > + struct bfin_serial_port *uart = (struct bfin_serial_port *)port;
> > + unsigned long flags;
> > + unsigned int baud, quot;
> > + unsigned short val, ier;
> > +
> > + baud = uart_get_baud_rate(port, termios, old, 0, port->uartclk/16);
> > + quot = bfin_serial_calc_baud(port->uartclk, baud);
> > + spin_lock_irqsave(&uart->port.lock, flags);
>
> Doesn't set port.ignore_status_mask nor port.read_status_mask, so *NONE*
> of the termios modes to do with error handling will work with this driver.
> In fact, I doubt it's possible to even change the parity, length or stop
> bit settings as this code stands.
Yes, not implemented, I'll do in future.
>
> > +
> > + /* Disable UART */
> > + ier = UART_GET_IER(uart);
> > + UART_PUT_IER(uart, 0);
> > +
> > + /* Set DLAB in LCR to Access DLL and DLH */
> > + val = UART_GET_LCR(uart);
> > + val |= DLAB;
> > + UART_PUT_LCR(uart, val);
> > + __builtin_bfin_ssync();
> > +
> > + UART_PUT_DLL(uart, quot & 0xFF);
> > + __builtin_bfin_ssync();
> > + UART_PUT_DLH(uart, (quot >> 8) & 0xFF);
> > + __builtin_bfin_ssync();
> > +
> > + /* Clear DLAB in LCR to Access THR RBR IER */
> > + val = UART_GET_LCR(uart);
> > + val &= ~DLAB;
> > + UART_PUT_LCR(uart, val);
> > + __builtin_bfin_ssync();
> > +
> > + /* Enable UART */
> > + UART_PUT_IER(uart, ier);
> > +
> > + val = UART_GET_GCTL(uart);
> > + val |= UCEN;
> > + UART_PUT_GCTL(uart, val);
> > +
> > + spin_unlock_irqrestore(&uart->port.lock, flags);
> > +}
> ...
> > +static int bfin_serial_calc_baud(unsigned int uartclk, unsigned int baud)
> > +{
> > + int quot;
> > +
> > + quot = uartclk / (baud * 8);
> > + if ((quot & 0x1) == 1) {
> > + quot++;
> > + }
> > + return quot/2;
>
> Something wrong with the standard function (uart_get_divisor)? If so, what?
No problem, will change to use it.
>
> > +}
> ...
> > +#ifdef CONFIG_SERIAL_BFIN_CONSOLE
> > +/*
> > + * Interrupts are disabled on entering
> > + */
> > +static void
> > +bfin_serial_console_write(struct console *co, const char *s, unsigned int count)
> > +{
> > + struct bfin_serial_port *uart = &bfin_serial_ports[co->index];
> > + int flags = 0;
> > + unsigned short status, tmp;
> > + int i;
> > +
> > + spin_lock_irqsave(&uart->port.lock, flags);
> > + for (i = 0; i < count; i++) {
> > + do {
> > + status = UART_GET_LSR(uart);
> > + } while (!(status & THRE));
> > +
> > + tmp = UART_GET_LCR(uart);
> > + tmp &= ~DLAB;
> > + UART_PUT_LCR(uart, tmp);
> > +
> > + UART_PUT_CHAR(uart, s[i]);
> > + if (s[i] == '\n') {
> > + do {
> > + status = UART_GET_LSR(uart);
> > + } while(!(status & THRE));
> > + UART_PUT_CHAR(uart, '\r');
> > + }
>
> Use uart_console_write()
>
> > + }
> > + spin_unlock_irqrestore(&uart->port.lock, flags);
> > +
> > +}
> > +
> > +/*
> > + * If the port was already initialised (eg, by a boot loader),
> > + * try to determine the current setup.
> > + */
> > +static void __init
> > +bfin_serial_console_get_options(struct bfin_serial_port *uart, int *baud,
> > + int *parity, int *bits)
> > +{
> > + unsigned short status;
> > +
> > + status = UART_GET_IER(uart) & (ERBFI | ETBEI);
> > + if (status == (ERBFI | ETBEI)) {
> > + /* ok, the port was enabled */
> > + unsigned short lcr, val;
> > + unsigned short dlh, dll;
> > +
> > + lcr = UART_GET_LCR(uart);
> > +
> > + *parity = 'n';
> > + if (lcr & PEN) {
> > + if (lcr & EPS)
> > + *parity = 'e';
> > + else
> > + *parity = 'o';
> > + }
> > + switch (lcr & 0x03) {
> > + case 0: *bits = 5; break;
> > + case 1: *bits = 6; break;
> > + case 2: *bits = 7; break;
> > + case 3: *bits = 8; break;
> > + }
>
> You do have length and parity settings on this port, yet you don't
> have code in your set_termios method to set them from the termios...
That means termios can't be changed on the fly. it will be enabled in future.
>
> > + /* Set DLAB in LCR to Access DLL and DLH */
> > + val = UART_GET_LCR(uart);
> > + val |= DLAB;
> > + UART_PUT_LCR(uart, val);
> > +
> > + dll = UART_GET_DLL(uart);
> > + dlh = UART_GET_DLH(uart);
> > +
> > + /* Clear DLAB in LCR to Access THR RBR IER */
> > + val = UART_GET_LCR(uart);
> > + val &= ~DLAB;
> > + UART_PUT_LCR(uart, val);
> > +
> > + *baud = get_sclk() / (16*(dll | dlh << 8));
> > + }
> > + pr_debug("%s:baud = %d, parity = %c, bits= %d\n", __FUNCTION__, *baud, *parity, *bits);
> > +}
>
Thanks for your comment and suggestion.
I'll get back to you soon.
-Aubrey
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists