Skip to contentSkip to conversation

char: fix logging when chardev write fails

PATCH Applied Daniel P. Berrangé 3 participants 3 messages Sep 14, 2020 – Sep 22, 2020

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 char: fix logging when chardev write fails

The qemu_chr_write_buffer() method sends data to the chardev backend for writing, and then also writes to the log file. In case the chardev backend only writes part of the data buffer, we need to make sure we only log the same subset. qemu_chr_write_buffer() will be invoked again later to write the rest of the buffer.

In the case the chardev backend returns an error though, no further attempts to likely to be made to write the data. We must therefore write the entire buffer to the log immediately.

An example where this is important is with the socket backend. This will return -1 for all writes if no client is currently connected. We still wish to write data to the log file when no client is present though. This used to work because the chardev would return "len" to pretend it had written all data when no client is connected, but this changed to return an error in

commit 271094474b65de1ad7aaf729938de3d9b9d0d36f
Author: Dima Stepanov <[email protected]>
Date: Thu May 28 12:11:18 2020 +0300

char-socket: return -1 in case of disconnect during tcp_chr_write

and this broke the logging, resulting in all data being discarded when no client is present.

---
 chardev/char.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
chardev/char.c +13 -0
index 77e7ec814f..e30cd025f5 100644
@@ -138,7 +138,20 @@ static int qemu_chr_write_buffer(Chardev *s,
         }
     }
     if (*offset > 0) {
+        /*
+         * If some data was written by backend, we should
+         * only log what was actually written. This method
+         * may be invoked again to write the remaining
+         * method, thus we'll log the remainder at that time.
+         */
         qemu_chr_write_log(s, buf, *offset);
+    } else if (res < 0) {
+        /*
+         * If a fatal error was reported by the backend,
+         * assume this method won't be invoked again with
+         * this buffer, so log it all right away.
+         */
+        qemu_chr_write_log(s, buf, len);
     }
     qemu_mutex_unlock(&s->chr_write_lock);
 
Signature · 1 line
2.26.2
Reviewed-by on v1 1/1
#1 lore raw
Quoted 22 lines · Daniel P. Berrangé

The qemu_chr_write_buffer() method sends data to the chardev backend for writing, and then also writes to the log file. In case the chardev backend only writes part of the data buffer, we need to make sure we only log the same subset. qemu_chr_write_buffer() will be invoked again later to write the rest of the buffer.

In the case the chardev backend returns an error though, no further attempts to likely to be made to write the data. We must therefore write the entire buffer to the log immediately.

An example where this is important is with the socket backend. This will return -1 for all writes if no client is currently connected. We still wish to write data to the log file when no client is present though. This used to work because the chardev would return "len" to pretend it had written all data when no client is connected, but this changed to return an error in

commit 271094474b65de1ad7aaf729938de3d9b9d0d36f
Author: Dima Stepanov <[email protected]>
Date: Thu May 28 12:11:18 2020 +0300

Quoted 7 lines

char-socket: return -1 in case of disconnect during tcp_chr_write

and this broke the logging, resulting in all data being discarded when no client is present.

Quoted 32 lines
---
 chardev/char.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
chardev/char.c +13 -1
index 77e7ec814f..e30cd025f5 100644
@@ -138,7 +138,20 @@ static int qemu_chr_write_buffer(Chardev *s,
         }
     }
     if (*offset > 0) {
+        /*
+         * If some data was written by backend, we should
+         * only log what was actually written. This method
+         * may be invoked again to write the remaining
+         * method, thus we'll log the remainder at that time.
+         */
         qemu_chr_write_log(s, buf, *offset);
+    } else if (res < 0) {
+        /*
+         * If a fatal error was reported by the backend,
+         * assume this method won't be invoked again with
+         * this buffer, so log it all right away.
+         */
+        qemu_chr_write_log(s, buf, len);
     }
     qemu_mutex_unlock(&s->chr_write_lock);
 
--

2.26.2

#2 lore raw
In reply to Daniel P. Berrangé
Quoted 48 lines · Daniel P. Berrangé

The qemu_chr_write_buffer() method sends data to the chardev backend for writing, and then also writes to the log file. In case the chardev backend only writes part of the data buffer, we need to make sure we only log the same subset. qemu_chr_write_buffer() will be invoked again later to write the rest of the buffer.

In the case the chardev backend returns an error though, no further attempts to likely to be made to write the data. We must therefore write the entire buffer to the log immediately.

An example where this is important is with the socket backend. This will return -1 for all writes if no client is currently connected. We still wish to write data to the log file when no client is present though. This used to work because the chardev would return "len" to pretend it had written all data when no client is connected, but this changed to return an error in

commit 271094474b65de1ad7aaf729938de3d9b9d0d36f
Author: Dima Stepanov <[email protected]>
Date: Thu May 28 12:11:18 2020 +0300

char-socket: return -1 in case of disconnect during tcp_chr_write

and this broke the logging, resulting in all data being discarded when no client is present.

---
 chardev/char.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
chardev/char.c +6 -0
index 77e7ec814f..e30cd025f5 100644
@@ -138,7 +138,20 @@ static int qemu_chr_write_buffer(Chardev *s,
         }
     }
     if (*offset > 0) {
+        /*
+         * If some data was written by backend, we should
+         * only log what was actually written. This method
+         * may be invoked again to write the remaining
+         * method, thus we'll log the remainder at that time.
+         */
         qemu_chr_write_log(s, buf, *offset);

chardev/char.c static int qemu_chr_write_buffer(Chardev *s, · line 157

Quoted 10 lines
chardev/char.c +7 -0
+    } else if (res < 0) {
+        /*
+         * If a fatal error was reported by the backend,
+         * assume this method won't be invoked again with
+         * this buffer, so log it all right away.
+         */
+        qemu_chr_write_log(s, buf, len);
     }
     qemu_mutex_unlock(&s->chr_write_lock);
 

Queued, thanks.

Paolo