qemu-devel
[Top][All Lists]
Advanced

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

Re: [Qemu-devel] [PATCH] qga-win: fall back to hardcoded user and group


From: Chen Hanxiao
Subject: Re: [Qemu-devel] [PATCH] qga-win: fall back to hardcoded user and group names if LookupAccountSidW failed
Date: Thu, 26 Oct 2017 19:01:47 +0800 (CST)

At 2017-10-26 17:59:34, "Michael Roth" <address@hidden> wrote:
>Quoting Chen Hanxiao (2017-10-26 04:27:40)
>> 
>> 
>> At 2017-10-26 05:58:07, "Michael Roth" <address@hidden> wrote:
>> >Quoting Tomáš Golembiovský (2017-09-29 16:31:02)
>> >> On Fri, 29 Sep 2017 17:11:22 +0800
>> >> Chen Hanxiao <address@hidden> wrote:
>> >> 
>> >> > From: Chen Hanxiao <address@hidden>
>> >> > 
>> >> > On some of windows (win08 sp2),
>> >> > it doesn't work by calling LookupAccountSidW with
>> >> > well-known SIDs,
>> >> > We got an error:
>> >> >     error 997 overlapped I/O operation is in progress
>> >> > 
>> >> > But hardcoded names work.
>> >> > 
>> >> > This patch introduces a workaroud for this issue:
>> >> > if LookupAccountSidW failed, try hardcoded one.
>> >> > 
>> >> > Signed-off-by: Chen Hanxiao <address@hidden>
>> >> > ---
>> >> >  qga/vss-win32/install.cpp | 10 ++++++++--
>> >> >  1 file changed, 8 insertions(+), 2 deletions(-)
>> >> > 
>> >> 
>> >> I wonder if it's caused by qga itself or a race/conflict with some other
>> >> app. But still...
>> >> 
>> >> 
>> >> Reviewed-by: Tomáš Golembiovský <address@hidden>
>> >
>> >I think it might be getNameByStringSID()/LookupAccountSidW() reporting a
>> >stale GetLastError() value.
>> >
>> >Does this fix the issue?
>> 
>> Not exactly.
>> I tested your patch several times, it improved greatly.
>> But failed only one time,
>> got another error 1722.
>
>Hmm, was that error also from the getNameByStringSID() call?

I don't know how to trace qemu-ga-win,
but added some fprintf(stdout).

+    hr = getNameByStringSID(administratorsGroupSID, buffer, &bufferLen);
+    if (FAILED(hr)) {

I saw my logs inside if (FAILED(hr)) {

But it looks weird, as your patch already dealing with this senario.


Regards,
- Chen

>
>> 
>> Build with my fallback patch and your suggestion, 
>> installation work perfectly.
>
>Thanks for testing.
>
>> 

reply via email to

[Prev in Thread] Current Thread [Next in Thread]