RE: bash 3.2.9-11 here doc permission problem

>> I'm having problems with "here docs" after an upgrade. Simply going >> back to bash 3.1-6 fixed the problem, but I'd thought I'd post cygchecks >> and straces. >The bash release recently changed to honor the setting of TMPDIR That was it! My script used TMPDIR; changed to use MYTMPDIR instea

RE: How to run a cygwin command from the window scheduler

> in a W2K3 command prompt (DOS) window > > C:\cywin\bin> (which is the command prompt prompt) > > I type the command > > c:\cygwin\bin\bash -c "/local/adminscripts/test.sh" Type "mount" instead from your DOS prompt while in the c:\cygwin\bin directory. If you see a mount point o

Re: window command in bash

>In a shell script what is the proper way to pass parameters to a window >program? Parameters are passed normally but special characters have to be "quoted". The backslashes (\) are interpreted by the shell, so quote each of them by entering \\ - - - - - Appended by Scientific Atlanta, a

Re: cp problem w/ snap 04-18 from remote (solved)

> The inode numbers of both files are identical. But in theory, cp should > only see both files as the same, if the inode numbers as well as the > device numbers are identical. You seem to suffer from a very seldom > situation, that the volume serial numbers (which are used as device > numbers)

RE: cp problem w/ snap 04-18 from remote

> I get > > cp: `//svtreboot3/logs/customer_logs.cfg' and `./customer_logs.cfg' are the same file > > when copying from a remote to a local machine. > > strace output is in attached file cp_problem.txt, and I've also attached the cygcheck -svr output in the .out > file. I forgot to mention that .

Re: "Segmentation fault" CygWin tools with code Injection-MS Detours

I hope you're not hard-coding c:\cygwin! -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain information which is confidential, proprietary, privileged or otherwise protected by law. The information is solely i

bash not noticing when child done

I've upgraded my cygwin and bash yesterday. I'm puzzled by what I see and every attempt I've made at generating a smaller test case has failed. Anyway, one script prints a message that it's about to call another script. On return from the script, another message is printed. The called script not

Re: [ANNOUNCEMENT] Updated: ash-20040127-3

> Wait for 20040127-3 and bash-3.0-7 to show up on the mirror and install > these. Did you really mean bash-3.0-7? I only see an announcement for bash-3.0-6. -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain

Re: check_case:strict problem

> Ok. I duplicated it. It has to be a share as in //foo/bar, not, > /foo/bar/blaz. The next snapshot should fix it. It did! Thanks! -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain information which is conf

Re: check_case:strict problem

>>Along similar lines as the ls failure on shares with >>CYGWIN=check_case:strict, if /etc/passwd has a home directory set to a >>share, bash won't even start. > >WJFFM. There was a problem pre-snapshot but bash still started. Now I >get no errors. Really? You DID edit your /etc/passwd and chan

check_case:strict problem (was Re: ls returns bad file descriptor)

Along similar lines as the ls failure on shares with CYGWIN=check_case:strict, if /etc/passwd has a home directory set to a share, bash won't even start. -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain informat

Re: ls returns bad file descriptor

> Btw, this should be fixed in the next snapshot. It works in 20050614 snapshot, thanks! -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain information which is confidential, proprietary, privileged or otherwis

Re: ls returns bad file descriptor

> Try turning off strict case checking. That works, but now I'm not unix-like. I put the strict case back and then tried the case that Windows shows in the explorer, Explr_drivers5\Reboot_results2, and ls still failed. In any event, looks like we're closer to solving the problem. -- Ken Shaffe

Re: ls returns bad file descriptor

> Perhaps there's some problem/issue with the permissions on the root of > your share that's confusing bash? Does this also happen in a > subdirectory of the share? bash hasn't changed, only cygwin1.dll. It also happens in a subdirectory of the share. I had posted an strace output which I hope

DEADBEEF in strace

Is this DEADBEEF from windows something to worry about? Program name: C:\cygwin\bin\basename.exe (pid 1204, ppid 1596) App version: 1005.16, api: 0.128 DLL version: 1005.16, api: 0.128 DLL build:2005-04-25 20:26 OS version: Windows NT-5.0 Heap size:402653184 Date/Time:2005-05-13 08

Re: unknown windows error 64

> Was this a copy to a network share? That error is > "ERROR_NETNAME_DELETED". Yes, it was a copy from a local disk to an archive established on a share. Is there a right way to copy to a share? I've done either of the following: cp file //machine/share/directory or cd //machine/share

unknown windows error 64

I'm having a problem with cp returning "Permission denied". I managed to strace it and found that after a whole lot of successful block read/writes, an unknown windows error 64 shows up. A snippet of the trace follows. 79 42539116 [main] cp 1324 readv: 1024 = readv (3, 0x22D590, 1), errno 0

Re: Losing track of processes?

"Christopher Faylor" wrote: >>> > If you're not using the -17 (test) version of bash, try that. >>> >>> Wow! Much better! My scripts are still churning after 4 hours. >>> When will this be part of an offical cygwin drop? >>Well, it's already "official" > > FWIW, I've removed -17 from test status

"strings" gone?

What package is "strings" in now. I thought it used to be in sh_utils but see it's being replaced by coreutils. -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain information which is confidential, proprietary,

BASH_VERSION vs cygcheck

Should the version info provided by "echo $BASH_VERSION" match the bash version displayed by "cygcheck -cd"? If not, is there some way to verify that the bash I'm really running is the version displayed by cygcheck? Or, does cygcheck gets its version info from within the executable? -- Ken Shaffe

Re: Losing track of processes?

"Brian Dessent " wrote: > If you're not using the -17 (test) version of bash, try that. Wow! Much better! My scripts are still churning after 4 hours. When will this be part of an offical cygwin drop? -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - -

Losing track of processes?

I have a suite of scripts which process logs but get hung after two hours. My initial looking into it shows that cygwin ps command thinks the processes are present, but windows task manager doesn't see them at all. It's as if the parent wasn't informed that it's child died. Perhaps a wait system c

Re: Please test 2005-Apr-12 snapshot

Floppy access stopped. Thanks! "Corinna Vinschen " wrote: > The first patch is less important, it should just stop floppy access > when sync is called on Win2K boxes. -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may

Re: floppy disk being accessed but shouldn't

/usr/bin/sync still accesses the floppy even though never written to. Cygwin DLL version info: DLL version: 1.5.15 DLL epoch: 19 DLL bad signal mask: 19005 DLL old termios: 5 DLL malloc env: 28 API major: 0 API minor: 126 Shared

Re: floppy disk being accessed but shouldn't

The shortest test case is "sync". Apparently it does something now. But no data has been buffered to the floppy. My scripts have this command strewn throughout to get around other problems with files created by other scripts. -- Ken Shaffer On Mon, 11 Apr 2005, Larry Hall wrote: > There was a

Re: setup next button always grayed out

"Bob Smart" wrote: > launch setup with the "-5" command-line option (suppress MD5 > verification). this worked! thanks. useful for the FAQ? -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain information which

floppy disk being accessed but shouldn't

Since I upgraded to 1.5.14, I hear the floppy drive being accessed and yet my scripts, to my knowledge, don't do this. Under 1.5.12, it was very quiet. I notice in the registry older mounts. Would cleaning these up help? Anyone know what tools or under what conditions a floppy drive "seek home" i

setup next button always grayed out

Under what conditions does setup not enable the Next button? It appears to have successfully finished the checksum and provide a list of packages, but never "enables" the Next button. I'm trying to install locally via UNC path which is a share in case that makes a difference. -- Ken Shaffer

ls taking too long

If a directory contains a large number of files (I have >4) where most of them are named such that the first character is a "1" and you do an "ls x*" where only one of the files begin with x, the ls takes an inordinate amount of time, but going to a plain dos window and "dir e*" is really fast.

Re: Using a.out.h from Cygwin with Mingw gcc - was Re: Binary read on textmode mount

I now include a.out.h via #include "/usr/include/a.out.h" and don't use the -I /usr/include which burned me. Fortunately, a.out.h has no #includes within it. More comments within yours below. -- Ken Shaffer On Fri, 5 Nov 2004, Larry Hall wrote: > But how appropriate it is to use 'a.out.h' or

Re: Binary read on textmode mount

Now, now, it seems perfectly legitimate to write a windows program which can access files created by gcc to run under cygwin. The structure of the a.out shouldn't depend on whether I'm compiling under windows, cygwin, linux, solaris, whatever. I just need the a.out.h file for the system on which

Re: Binary read on textmode mount

I need a.out.h apparently not available with the -mno-cygwin compile option. -- Ken Shaffer On Fri, 5 Nov 2004, Christopher Faylor wrote: > You are really asking for trouble. The fact that you needed to go out > of your way to include a cygwin header file in a windows program should > be a clu

Re: Binary read on textmode mount

I see the problem now... O_BINARY is 0x8000 on mingw and 0x1 in cygwin and I needed a header file from cygwin and use the -I option which brought in the wrong fcntl.h Sigh. -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any att

Re: Binary read on textmode mount

I'm using the cygwin development environment and so thought this was the correct place to post a problem. But you make a good point, so I copied the file to /tmp, mounted as binary, and got the same results. So, I guess the gcc libraries are at fault? Is this the right list? -- Ken Shaffer On

flushing not always done on exit?

To solve some weird "periodic" hangs where a script depends on the output of a previous portion of the script, I've had to insert "sync" command. My scripts use awk, sed, and grep, to name a few. I haven't come up with a simple test case yet, but thought I'd raise a caution flag. -- Ken Shaffer

mv performing a cp?

If I attempt to rename a directory via mv and a file within that directory is in use, instead of saying the mv can't be done, mv tries to cp. Is this expected behaviour? Prior to mv: /c/logs$ ls -ld kimberly drwxrwxrwx+ 181 Administ Domain U 327680 Oct 20 10:58 kimberly /c/logs$ I happen to h

Re: setup 2.427 runtime error

David A. Cobb <[EMAIL PROTECTED]> > To: [EMAIL PROTECTED], [EMAIL PROTECTED] > Cc: Cygwin Application Packaging Discussion List <[EMAIL PROTECTED]> > Subject: Re: setup 2.427 runtime error > > Shaffer, Kenneth wrote: > > >I'm getting a setup.exe runtime er

setup 2.427 runtime error

I'm getting a setup.exe runtime error when trying to install from a local disk (after downloading from the internet) just after the MD5 checks. I thought it was about time I upgrade. My current cygwin1.dll version 1.5.5 (does setup use the dll?) -- Ken Shaffer - - - - - - - Appended by

ps command showing unknown flags

I recently ran the ps command and saw an unknown flag, "I", displayed in column 1. There was no heading above it and the man page wasn't much help. What does this mean? Where can I find documentation on perhaps other flags? -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, I

Re: terminal size problems

rlogin, to my knowledge, has nothing to do with the X protocol of resizing. After you resize your window, set some variables accordingly: LINES=height COLUMNS=width export LINES COLUMNS If you start with an xterm, you shouldn't have a problem, but since you started with a console window and ju

questions on signals

How do I get trap to get back to the default settings after entering 'trap "echo hi" DEBUG'? 'trap -' doesn't seem to work. How can I get ctrl-C to quit this cpu-intensive loop? echo | awk '{while(1){} }' cygcheck shows: bash2.05b-16 OK gawk

Re: Executing Perl scripts w/o the .pl

If using Active Perl with cygwin, put the following as the first line of your perl scripts: eval 'exec perl -S `cygpath -w $0`' if 0; -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain information which is confid

Re: perl localtime returns gmttime

I haven't seen closure on this subject yet. I too have the same problem. -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain information which is confidential, proprietary, privileged or otherwise protected by l

Re: cvs complains of no repository (solved)

this in the CVS FAQ if there is one. -- Ken Shaffer On Thu, 13 Nov 2003, Shaffer, Kenneth wrote: > I looked further into this and think the problem is due to the fact > that the 1.10 version I was using was strictly Windows-based and so > whenever a CVS/Root file was created it had the

Re: cvs complains of no repository

I looked further into this and think the problem is due to the fact that the 1.10 version I was using was strictly Windows-based and so whenever a CVS/Root file was created it had the for the end of line. Then when I used the cygwin version I had trouble. The following patch fixed the problem I w

cvs complains of no repository

There's a difference between 1.10 and 1.11 that I don't understand. Performing a "cvs co -c" using 1.10 correctly provides a listing of the projects. The version shows "Concurrent Versions System (CVS) 1.10 `Halibut' (client)". Performing a "cvs co -c" using 1.11.6 just says ": no such repository

Bad interpreter?

If the first line of scripts is #!bash.exe, I get back a bad interpreter message. Is there a way that I can get the above line to correctly find bash? I thought if the full path wasn't specified, that PATH would be used to find it. (I've tried copying bash.exe to / and c: and get the same error m

where is new stdint.h?

I'm wondering if the new stdint.h is available and if so, what package would it be in, Devel? -- Ken Shaffer - - - - - - - Appended by Scientific-Atlanta, Inc. - - - - - - - This e-mail and any attachments may contain information which is confidential, proprietary, privileged or other

Re: iconv.h

> Where/how I can get iconv.h? Feel free to post to me directly, > as well as the list. It's probably in the Devel category as listed by setup which by default isn't installed. -- Ken Shaffer Staff Software Engineer 770-236-3421 [EMAIL PROTECTED] - - - - - - - Appended by Scientific-Atla