When compiled with x86_64-w64-mingw32, we have:
error: implicit declaration of function 'qemu_log_mask'
error: 'LOG_UNIMP' undeclared (first use in this function)
This patch adds the missing include.
Fixes: 7299e1a411
("hw/i386/vmport: replace fprintf() by trace events or LOG_UNIMP")
Signed-off-by: Laurent Vivier <laurent@vivier.eu>
---
hw/i386/vmport.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/hw/i386/vmport.c b/hw/i386/vmport.c
index 9b8c68806e..116aa09819 100644
--- a/hw/i386/vmport.c
+++ b/hw/i386/vmport.c
@@ -27,6 +27,7 @@
#include "hw/i386/pc.h"
#include "sysemu/hw_accel.h"
#include "hw/qdev.h"
+#include "qemu/log.h"
#include "trace.h"
#define VMPORT_CMD_GETVERSION 0x0a
--
2.14.3
On 21 December 2017 at 21:11, Laurent Vivier <laurent@vivier.eu> wrote: > When compiled with x86_64-w64-mingw32, we have: > > error: implicit declaration of function 'qemu_log_mask' > error: 'LOG_UNIMP' undeclared (first use in this function) > > This patch adds the missing include. > > Fixes: 7299e1a411 > ("hw/i386/vmport: replace fprintf() by trace events or LOG_UNIMP") > Signed-off-by: Laurent Vivier <laurent@vivier.eu> Looks to me like it'll be a bug for anything not configured using the 'log' trace backend, rather than win32 specific. I'll apply (and fix up the commit message). thanks -- PMM
>> Fixes: 7299e1a411 >> ("hw/i386/vmport: replace fprintf() by trace events or LOG_UNIMP") >> Signed-off-by: Laurent Vivier <laurent@vivier.eu> > > Looks to me like it'll be a bug for anything not configured > using the 'log' trace backend, rather than win32 specific. Sorry my bad, it seems I don't test enough lately, I'll try to do more testing, and somehow to reduce CI build time, more they last, less we rely on it, even with personal repositories.
On 21 December 2017 at 22:55, Philippe Mathieu-Daudé <f4bug@amsat.org> wrote: >>> Fixes: 7299e1a411 >>> ("hw/i386/vmport: replace fprintf() by trace events or LOG_UNIMP") >>> Signed-off-by: Laurent Vivier <laurent@vivier.eu> >> >> Looks to me like it'll be a bug for anything not configured >> using the 'log' trace backend, rather than win32 specific. > > Sorry my bad, it seems I don't test enough lately, I'll try to do more > testing, and somehow to reduce CI build time, more they last, less we > rely on it, even with personal repositories. It's unfortunate that trace.h pulls in qemu/log.h if you're using the log backend, because it makes this particular mistake easy; we've had quite a few instances of it... (I could probably configure one of my test setups with a different trace backend which would at least catch these before they hit master.) thanks -- PMM
© 2016 - 2024 Red Hat, Inc.