Patchwork linux-user/strace.c: Correct errno printing for mmap etc

login
register
mail settings
Submitter Peter Maydell
Date Nov. 21, 2011, 12:21 p.m.
Message ID <1321878079-30836-1-git-send-email-peter.maydell@linaro.org>
Download mbox | patch
Permalink /patch/126772/
State New
Headers show

Comments

Peter Maydell - Nov. 21, 2011, 12:21 p.m.
Correct the printing of errnos for syscalls which are handled
via print_syscall_ret_addr (mmap, mmap2, brk, shmat): errnos
are returned as negative returned values at this level, not
via the host 'errno' variable.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
This applies on top of Alex's [v3] linux-user: fix QEMU_STRACE=1 segfault
patch. It is fixing a separate bug to that patch, but OTOH it does touch
only four lines of actual code, all of which were added in that patch.
Keep it separate or fold it in with that one, I don't mind.

 linux-user/strace.c |    9 ++++-----
 1 files changed, 4 insertions(+), 5 deletions(-)
Riku Voipio - Nov. 23, 2011, 1:15 p.m.
On Mon, Nov 21, 2011 at 12:21:19PM +0000, Peter Maydell wrote:
> Correct the printing of errnos for syscalls which are handled
> via print_syscall_ret_addr (mmap, mmap2, brk, shmat): errnos
> are returned as negative returned values at this level, not
> via the host 'errno' variable.
 
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
> This applies on top of Alex's [v3] linux-user: fix QEMU_STRACE=1 segfault
> patch. It is fixing a separate bug to that patch, but OTOH it does touch
> only four lines of actual code, all of which were added in that patch.
> Keep it separate or fold it in with that one, I don't mind.

I think in general it is recommended to squash patches that only change lines
added in the previous patch. Otoh it muddies the authorship of the change.
Thus, unless someone objects, I'll que these as separate patches so both
of you get clear credit.

Riku

>  linux-user/strace.c |    9 ++++-----
>  1 files changed, 4 insertions(+), 5 deletions(-)
> 
> diff --git a/linux-user/strace.c b/linux-user/strace.c
> index 269481e..05a0d3e 100644
> --- a/linux-user/strace.c
> +++ b/linux-user/strace.c
> @@ -1,5 +1,4 @@
>  #include <stdio.h>
> -#include <errno.h>
>  #include <sys/ipc.h>
>  #include <sys/msg.h>
>  #include <sys/sem.h>
> @@ -286,11 +285,11 @@ print_syscall_ret_addr(const struct syscallname *name, abi_long ret)
>  {
>      char *errstr = NULL;
>  
> -    if (ret == -1) {
> -        errstr = target_strerror(errno);
> +    if (ret < 0) {
> +        errstr = target_strerror(-ret);
>      }
> -    if ((ret == -1) && errstr) {
> -        gemu_log(" = -1 errno=%d (%s)\n", errno, errstr);
> +    if (errstr) {
> +        gemu_log(" = -1 errno=%d (%s)\n", (int)-ret, errstr);
>      } else {
>          gemu_log(" = 0x" TARGET_ABI_FMT_lx "\n", ret);
>      }
> -- 
> 1.7.1
Alexander Graf - Nov. 23, 2011, 1:18 p.m.
On 11/23/2011 02:15 PM, Riku Voipio wrote:
> On Mon, Nov 21, 2011 at 12:21:19PM +0000, Peter Maydell wrote:
>> Correct the printing of errnos for syscalls which are handled
>> via print_syscall_ret_addr (mmap, mmap2, brk, shmat): errnos
>> are returned as negative returned values at this level, not
>> via the host 'errno' variable.
>
>> Signed-off-by: Peter Maydell<peter.maydell@linaro.org>
>> ---
>> This applies on top of Alex's [v3] linux-user: fix QEMU_STRACE=1 segfault
>> patch. It is fixing a separate bug to that patch, but OTOH it does touch
>> only four lines of actual code, all of which were added in that patch.
>> Keep it separate or fold it in with that one, I don't mind.
> I think in general it is recommended to squash patches that only change lines
> added in the previous patch. Otoh it muddies the authorship of the change.
> Thus, unless someone objects, I'll que these as separate patches so both
> of you get clear credit.

I am completely indifferent :).


Alex

Patch

diff --git a/linux-user/strace.c b/linux-user/strace.c
index 269481e..05a0d3e 100644
--- a/linux-user/strace.c
+++ b/linux-user/strace.c
@@ -1,5 +1,4 @@ 
 #include <stdio.h>
-#include <errno.h>
 #include <sys/ipc.h>
 #include <sys/msg.h>
 #include <sys/sem.h>
@@ -286,11 +285,11 @@  print_syscall_ret_addr(const struct syscallname *name, abi_long ret)
 {
     char *errstr = NULL;
 
-    if (ret == -1) {
-        errstr = target_strerror(errno);
+    if (ret < 0) {
+        errstr = target_strerror(-ret);
     }
-    if ((ret == -1) && errstr) {
-        gemu_log(" = -1 errno=%d (%s)\n", errno, errstr);
+    if (errstr) {
+        gemu_log(" = -1 errno=%d (%s)\n", (int)-ret, errstr);
     } else {
         gemu_log(" = 0x" TARGET_ABI_FMT_lx "\n", ret);
     }