[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v8 13/26] emul/ns16550: implement dump_state() hook





On 11/11/26 8:47 PM, dmukhin@xxxxxxxx wrote:
From: Denis Mukhin <dmukhin@xxxxxxxx>

At the commit subject:

Nit: emul/ns16x50: like the rest of the series.


Implement dump_state() vUART hook for debugging UART state machine over
Xen console. dump_state() prints state of all emulated registers
(including state-less IIR) and state of RX/TX FIFOs.

Signed-off-by: Denis Mukhin <dmukhin@xxxxxxxx>
---
  xen/common/emul/vuart/ns16x50.c | 64 ++++++++++++++++++++++++++++++++-
  1 file changed, 63 insertions(+), 1 deletion(-)

diff --git a/xen/common/emul/vuart/ns16x50.c b/xen/common/emul/vuart/ns16x50.c
index 404e12311090..cda21c15cf78 100644
--- a/xen/common/emul/vuart/ns16x50.c
+++ b/xen/common/emul/vuart/ns16x50.c
@@ -728,6 +728,62 @@ static int ns16x50_io_read(
      return rc;
  }
+static void cf_check ns16x50_dump_state(void *arg)
+{
+#ifdef CONFIG_VUART_NS16X50_DEBUG

With the whole body under CONFIG_VUART_NS16X50_DEBUG, the 'q'
keyhandler prints nothing for the vUART on a normal build, which is
where a state dump is most useful. The function doesn't cost anything
until someone presses 'q'; I'd drop the #ifdef (or, if it must stay,
leave .dump_state NULL instead of registering an empty hook).

+    struct vuart_ns16x50 *vdev = arg;
+    const struct domain *d;
+    const struct vuart_info *info;
+    const struct xencons_interface *cons;
+    const uint8_t *regs;
+    const char *as = (info->base_type & IORESOURCE_IO) ? "ioport":
+                                                         "iomem";

As mentioned in my reply to the cover letter, info is used before it
is assigned; gcc rejects this with -Werror=uninitialized, so the series
doesn't build with CONFIG_VUART_NS16X50_DEBUG=y. Computing as after
info = vdev->info fixes it.

+    unsigned long flags;
+
+    ASSERT(vdev);
+
+    /* Allow printing state in case of a deadlock. */
+    if ( !ns16x50_trylock(vdev, &flags) )
+        return;
+
+    d = vdev->owner;
+    info = vdev->info;
+    cons = &vdev->cons;
+    regs = &vdev->regs[0];
+
+    printk("Virtual " pr_prefix " (%s) %s %#"PRIx64" IRQ#%d owner %pd\n",
+            vdev->name, as, info->base_addr, info->irq, d);

Nits on the format strings: PRIpaddr for base_addr, %u for irq and the
XENCONS_RING_IDX values, %zu for ARRAY_SIZE(), and PRIx8 for UART_RBR
and friends, which are plain ints. Also the continuation lines are
indented one column too far.

+
+    printk("  RX FIFO size %ld in_prod %d in_cons %d used %d\n",
+            ARRAY_SIZE(cons->in), cons->in_prod, cons->in_cons,
+            cons->in_prod - cons->in_cons);
+
+    printk("  TX FIFO size %ld out_prod %d out_cons %d used %d\n",
+            ARRAY_SIZE(cons->out), cons->out_prod, cons->out_cons,
+            cons->out_prod - cons->out_cons);
+
+    printk("  %02"PRIx8" RBR %02"PRIx8" THR %02"PRIx8" DLL %02"PRIx8" DLM 
%02"PRIx8"\n",
+            UART_RBR,
+            cons->in[MASK_XENCONS_IDX(cons->in_prod, cons->in)],
+            cons->out[MASK_XENCONS_IDX(cons->out_prod, cons->out)],
+            regs[NS16X50_REGS_NUM + UART_DLL],
+            regs[NS16X50_REGS_NUM + UART_DLM]);

in_prod and out_prod point at the next free slot, so these print
whatever stale byte happens to be there. For RBR, the character the
guest will read next is in[in_cons] (only meaningful if the FIFO isn't
empty); for THR, the last character written is out[out_prod - 1].

+
+    printk("  %02"PRIx8" IER %02"PRIx8"\n", UART_IER, regs[UART_IER]);
+
+    printk("  %02"PRIx8" FCR %02"PRIx8" IIR %02"PRIx8"\n",
+            UART_FCR, regs[UART_FCR], ns16x50_iir_get(vdev));
+
+    printk("  %02"PRIx8" LCR %02"PRIx8"\n", UART_LCR, regs[UART_LCR]);
+    printk("  %02"PRIx8" MCR %02"PRIx8"\n", UART_MCR, regs[UART_MCR]);
+    printk("  %02"PRIx8" LSR %02"PRIx8"\n", UART_LSR, regs[UART_LSR]);
+    printk("  %02"PRIx8" MSR %02"PRIx8"\n", UART_MSR, regs[UART_MSR]);
+    printk("  %02"PRIx8" SCR %02"PRIx8"\n", UART_SCR, regs[UART_SCR]);
+
+    ns16x50_unlock(vdev, &flags);
+#endif /* CONFIG_VUART_NS16X50_DEBUG */
+}
+
  static void ns16x50_access(int dir, unsigned int type,
                             unsigned long addr, unsigned int size,
                             uint32_t *data)
@@ -794,6 +850,9 @@ static void ns16x50_access(int dir, unsigned int type,
          rc = ns16x50_io_read(vdev, reg, size, data);
+ if ( ns16x50_log_level >= 3 )
+        ns16x50_dump_state(vdev);

ns16x50_log_level is at most 2, so this never runs (see 04/26). If it
did, it would run with vdev->lock held, so the trylock in
ns16x50_dump_state() would fail and print nothing. I'd drop both
call sites added here and keep dump_state() for the keyhandler only. Another call site is here ...

+
      if ( rc == 0 )
          ns16x50_debug(vdev, "%c %s %#"PRIx64" %d: LCR=%#02x %#02x %#08x\n",
                        op, as, addr, size, vdev->regs[UART_LCR], reg, *data);
@@ -971,6 +1030,9 @@ static int cf_check ns16x50_put_rx(void *arg, char ch)
ns16x50_unlock(vdev, &flags); + if ( ns16x50_log_level >= 3 )
+        ns16x50_dump_state(vdev);

... ^.

+
      return rc;
  }
@@ -979,7 +1041,7 @@ static int cf_check ns16x50_put_rx(void *arg, char ch)
      .compatible = "ns16550",            \
      .alloc      = ns16x50_alloc,        \
      .free       = ns16x50_free,         \
-    .dump_state = NULL,                 \
+    .dump_state = ns16x50_dump_state,   \
      .put_rx     = ns16x50_put_rx,       \
  }

~ Oleksii




 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.