Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
Hi Chris, On Jan 11 10:58, Christopher Faylor wrote: On Mon, Jan 07, 2008 at 08:12:32AM -0500, Christopher Faylor wrote: On Mon, Jan 07, 2008 at 02:11:02PM +0100, Corinna Vinschen wrote: On Dec 23 10:28, Corinna Vinschen wrote: On Dec 22 11:08, Christopher Faylor wrote: I have some pipe-related patches that probably should go into the branch too, FYI. I'm just waiting for word that they fix a reported problem. Ok, I'll wait. No worries. Did you get feedback about your change? No. I had some email problems in the last month so I'm resyncing with the people who would be testing. I'm waiting for feedback now... Any news? Out of curiosity, would you mind to send a sneak preview of the patch and what problem it's meant to solve? Thanks, Corinna -- Corinna Vinschen Please, send mails regarding Cygwin to Cygwin Project Co-Leader cygwin AT cygwin DOT com Red Hat
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
On Wed, Jan 16, 2008 at 12:19:39PM +0100, Corinna Vinschen wrote: On Jan 11 10:58, Christopher Faylor wrote: On Mon, Jan 07, 2008 at 08:12:32AM -0500, Christopher Faylor wrote: On Mon, Jan 07, 2008 at 02:11:02PM +0100, Corinna Vinschen wrote: On Dec 23 10:28, Corinna Vinschen wrote: On Dec 22 11:08, Christopher Faylor wrote: I have some pipe-related patches that probably should go into the branch too, FYI. I'm just waiting for word that they fix a reported problem. Ok, I'll wait. No worries. Did you get feedback about your change? No. I had some email problems in the last month so I'm resyncing with the people who would be testing. I'm waiting for feedback now... Any news? No. Out of curiosity, would you mind to send a sneak preview of the patch and what problem it's meant to solve? The sneak preview is already in 1.7.x. There is a problem with the code in 1.7.x (as you know) which shouldn't be an issue for 1.5.x. cgf
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
On Jan 16 09:50, Christopher Faylor wrote: On Wed, Jan 16, 2008 at 12:19:39PM +0100, Corinna Vinschen wrote: On Jan 11 10:58, Christopher Faylor wrote: On Mon, Jan 07, 2008 at 08:12:32AM -0500, Christopher Faylor wrote: On Mon, Jan 07, 2008 at 02:11:02PM +0100, Corinna Vinschen wrote: On Dec 23 10:28, Corinna Vinschen wrote: On Dec 22 11:08, Christopher Faylor wrote: I have some pipe-related patches that probably should go into the branch too, FYI. I'm just waiting for word that they fix a reported problem. Ok, I'll wait. No worries. Did you get feedback about your change? No. I had some email problems in the last month so I'm resyncing with the people who would be testing. I'm waiting for feedback now... Any news? No. Out of curiosity, would you mind to send a sneak preview of the patch and what problem it's meant to solve? The sneak preview is already in 1.7.x. There is a problem with the code in 1.7.x (as you know) which shouldn't be an issue for 1.5.x. Uh, ok. Thanks, Corinna -- Corinna Vinschen Please, send mails regarding Cygwin to Cygwin Project Co-Leader cygwin AT cygwin DOT com Red Hat
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
On Mon, Jan 07, 2008 at 08:12:32AM -0500, Christopher Faylor wrote: On Mon, Jan 07, 2008 at 02:11:02PM +0100, Corinna Vinschen wrote: Hi Chris, On Dec 23 10:28, Corinna Vinschen wrote: On Dec 22 11:08, Christopher Faylor wrote: On Sat, Dec 22, 2007 at 10:48:51AM +0100, Corinna Vinschen wrote: On Dec 20 19:35, Brian Dessent wrote: Christopher Faylor wrote: Unless Corinna says differently, I think she wants to be in control of what goes into the branch so I don't want to suggest that you should check it in there too. Okay, I'll let her take care of the branch since she's been handling all the releases from it. Please check it in to the branch. I guess it might be a good idea to release a 1.5.27-8 at one point. I'll be more or less unavailable the next two weeks but perhaps I can release a new version over christmas. I have some pipe-related patches that probably should go into the branch too, FYI. I'm just waiting for word that they fix a reported problem. Ok, I'll wait. No worries. Did you get feedback about your change? No. I had some email problems in the last month so I'm resyncing with the people who would be testing. I'm waiting for feedback now... cgf
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
Hi Chris, On Dec 23 10:28, Corinna Vinschen wrote: On Dec 22 11:08, Christopher Faylor wrote: On Sat, Dec 22, 2007 at 10:48:51AM +0100, Corinna Vinschen wrote: On Dec 20 19:35, Brian Dessent wrote: Christopher Faylor wrote: Unless Corinna says differently, I think she wants to be in control of what goes into the branch so I don't want to suggest that you should check it in there too. Okay, I'll let her take care of the branch since she's been handling all the releases from it. Please check it in to the branch. I guess it might be a good idea to release a 1.5.27-8 at one point. I'll be more or less unavailable the next two weeks but perhaps I can release a new version over christmas. I have some pipe-related patches that probably should go into the branch too, FYI. I'm just waiting for word that they fix a reported problem. Ok, I'll wait. No worries. Did you get feedback about your change? Corinna -- Corinna Vinschen Please, send mails regarding Cygwin to Cygwin Project Co-Leader cygwin AT cygwin DOT com Red Hat
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
On Mon, Jan 07, 2008 at 02:11:02PM +0100, Corinna Vinschen wrote: Hi Chris, On Dec 23 10:28, Corinna Vinschen wrote: On Dec 22 11:08, Christopher Faylor wrote: On Sat, Dec 22, 2007 at 10:48:51AM +0100, Corinna Vinschen wrote: On Dec 20 19:35, Brian Dessent wrote: Christopher Faylor wrote: Unless Corinna says differently, I think she wants to be in control of what goes into the branch so I don't want to suggest that you should check it in there too. Okay, I'll let her take care of the branch since she's been handling all the releases from it. Please check it in to the branch. I guess it might be a good idea to release a 1.5.27-8 at one point. I'll be more or less unavailable the next two weeks but perhaps I can release a new version over christmas. I have some pipe-related patches that probably should go into the branch too, FYI. I'm just waiting for word that they fix a reported problem. Ok, I'll wait. No worries. Did you get feedback about your change? No. I had some email problems in the last month so I'm resyncing with the people who would be testing. cgf
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
Brian Dessent wrote: ... but sadly a similar cleanup is still required for cygpath as well if we are to support 9x/ME out of the 1.5 branch. In that case I suppose you could just revert cygpath.cc to an older revision before the native APIs were added. Er, nevermind. I was accidently looking at HEAD, but the native stuff in cygpath is not on the branch. So I think only the bloda.cc change is necessary to restore 9x/ME capability on the branch. Brian
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
On Thu, Dec 20, 2007 at 07:15:53AM -0800, Brian Dessent wrote: Brian Dessent wrote: Fortunately, I have VMware with a Win98 image here. The problem is that bloda.c calls NtQuerySystemInformation without using any kind of autoload.cc-type indirection, and so cygcheck gets a hard dependency on ntdll.dll which doesn't exist on 9x/ME. We need to do one of: - Revert the bloda-check feature on the 1.5 branch - Check windows version at runtime and only do NT calls through LoadLibrary/GetProcAddress - Use the autoload.cc trick in cygcheck If we're going to make releases from the 1.5 branch then I don't think it's quite acceptible just yet to shaft 9x users, after all that's the whole point of the branch. The attached patch un-NT-ifies bloda.cc but sadly a similar cleanup is still required for cygpath as well if we are to support 9x/ME out of the 1.5 branch. In that case I suppose you could just revert cygpath.cc to an older revision before the native APIs were added. Brian 2007-12-20 Brian Dessent [EMAIL PROTECTED] * Makefile.in (cygcheck.exe): Don't link to ntdll. * bloda.cc (pNtQuerySystemInformation): Add. (pRtlAnsiStringToUnicodeString): Add. (get_process_list): Use function pointers for NT functions. (dump_dodgy_apps): Skip dodgy app check on non-NT platforms. Use GetProcAddress for NT-specific functions. I had something similar in my sandbox but 1) you beat me to it and 2) yours is better. So, please check this into the trunk. I don't have any problem with cygcheck being Windows 9x aware since it could help us track down problems with people who are trying to run Cygwin 1.7.x on older systems. Unless Corinna says differently, I think she wants to be in control of what goes into the branch so I don't want to suggest that you should check it in there too. cgf
RE: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
On 20 December 2007 21:12, Christopher Faylor wrote: On Thu, Dec 20, 2007 at 07:15:53AM -0800, Brian Dessent wrote: The attached patch un-NT-ifies bloda.cc but sadly a similar cleanup is still required for cygpath as well if we are to support 9x/ME out of the 1.5 branch. In that case I suppose you could just revert cygpath.cc to an older revision before the native APIs were added. I had something similar in my sandbox but 1) you beat me to it and 2) yours is better. So, please check this into the trunk. I don't have any problem with cygcheck being Windows 9x aware since it could help us track down problems with people who are trying to run Cygwin 1.7.x on older systems. Unless Corinna says differently, I think she wants to be in control of what goes into the branch so I don't want to suggest that you should check it in there too. But it only belongs on the branch at all, doesn't it? When I wrote bloda.cc, I didn't bother with 9x compat, because I was only writing it for HEAD, where we have stopped supporting 9x. Surely there are many other reasons why HEAD won't work on 9x, so the only benefit would be in applying this patch to the branch? cheers, DaveK -- Can't think of a witty .sigline today
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
On Fri, Dec 21, 2007 at 01:59:54AM -, Dave Korn wrote: On 20 December 2007 21:12, Christopher Faylor wrote: On Thu, Dec 20, 2007 at 07:15:53AM -0800, Brian Dessent wrote: The attached patch un-NT-ifies bloda.cc but sadly a similar cleanup is still required for cygpath as well if we are to support 9x/ME out of the 1.5 branch. In that case I suppose you could just revert cygpath.cc to an older revision before the native APIs were added. I had something similar in my sandbox but 1) you beat me to it and 2) yours is better. So, please check this into the trunk. I don't have any problem with cygcheck being Windows 9x aware since it could help us track down problems with people who are trying to run Cygwin 1.7.x on older systems. Unless Corinna says differently, I think she wants to be in control of what goes into the branch so I don't want to suggest that you should check it in there too. But it only belongs on the branch at all, doesn't it? When I wrote bloda.cc, I didn't bother with 9x compat, because I was only writing it for HEAD, where we have stopped supporting 9x. Surely there are many other reasons why HEAD won't work on 9x, so the only benefit would be in applying this patch to the branch? I explained my logic above: I don't have any problem with cygcheck being Windows 9x aware since it could help us track down problems with people who are trying to run Cygwin 1.7.x on older systems. The problem in this case would be Hey! Look at what cygcheck is saying! You are using Windows 9x! You can't do that! OTOH, you could make a case that cygcheck on the trunk should just consider Windows 9x and friends to be dodgy apps and should issue a clear error in that case. cgf
Re: [patch] un-NT-ify cygcheck (was: cygwin 1.5.25-7: cygcheck does not work?)
Christopher Faylor wrote:
Unless Corinna says differently, I think she wants to be in control of
what goes into the branch so I don't want to suggest that you should
check it in there too.
Okay, I'll let her take care of the branch since she's been handling all
the releases from it.
The problem in this case would be Hey! Look at what cygcheck is saying!
You are using Windows 9x! You can't do that!
In a sense it already does this:
case VER_PLATFORM_WIN32_WINDOWS:
switch (osversion.dwMinorVersion)
{
case 0:
osname = 95 (not supported);
break;
case 10:
osname = 98 (not supported);
break;
case 90:
osname = ME (not supported);
break;
default:
osname = 9X (not supported);
break;
}
break;
Brian
