September 2008 - Darwin-dev

AppleUSBIrDA Released
by William Siegrist 16 Sep '08

16 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com http://www.opensource.apple.com/darwinsource/10.5/ Thank you. -- William Siegrist Darwin Team _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com The Darwin Team is pleased to announce the release of AppleUSBIrDA-144.4.0 which is the version for all Leopard builds: smime.p7s

1 0

spinlock under fork()?
by Nathan Herring 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thread-index: AckXjwEwfNv0PUmwSi+LuOm6uij5sA== Thread-topic: spinlock under fork()? _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: GNU assembler freaks out at jumps and Intel syntax
by LuigiG 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=X-YMail-OSG:Received:X-Mailer:Date:From:Reply-To:Subject:To:In-Reply-To:MIME-Version:Content-Type:Message-ID; b=n+5vNscuoRWN9JprMudbNvGvI2nkWtIIxDzHRUMn29xE1mPcJePvY99dV0i8G/z4mBvO90WMoKTDtP/1vVMY/+qpfThoahH67Bsv0mrRaH3vcqolushpkemPB1PQbdYOokxCtnyj3YuFhAG2mskoL2gfv9++BgpQ4feOWMq7aaA=; Not quite sure what you mean with that, is the AT&T syntax more "logical" for any particular reason? Anyway, my reason is a compiler emitting asm (Intel syntax), I'd planned to use gas a back-end. It needs to able to address labels at any arbitrary position in the source, therefore the 1b/1f isn't good enough. Though I cannot find any specific reference in the docs concering use of "j<cc>" instructions when using intel syntax ,this is starting to look like a bug in gas itself. I'll give a shot at the binutils mailing list. Thanks anyway. Luigi --- On Tue, 9/16/08, Terry Lambert <tlambert(a)apple.com> wrote: > From: Terry Lambert <tlambert(a)apple.com> > Subject: Re: GNU assembler freaks out at jumps and Intel syntax > To: news4e71(a)yahoo.com > Cc: darwin-dev(a)lists.apple.com > Date: Tuesday, September 16, 2008, 1:26 AM > Ah, you have some reason other than logic for using Intel > syntax. > > Change it to this: > > 1: > L5: > popl %ebx > cmpl $2, 8(%ebp) > je L2 > leal LC0-"L00000000001$pb"(%ebx), %eax > movl %eax, (%esp) > call L_puts$stub > .intel_syntax noprefix > test %eax, %eax > jnz 1b > .att_syntax > > -- Terry > > > On Sep 15, 2008, at 4:15 PM, LuigiG wrote: > > But that's using the AT&T syntax. That has > always worked fine for me > > too, it's only when you switch to Intel syntax > that you get problems. > > > > If I chnage your example to use Intel syntax (even > limited to that > > small bit of code), it doesn't work anymore: > > > > This assembles (and links) correctly: > > testl %eax,%eax > > jnz L5 > > > > This will fail: > > .intel_syntax noprefix > > test eax,eax > > jnz L5 > > .att_syntax > > > > foo.s:28:suffix or operands invalid for `jnz' > > > > So, is this a bug in gas? Or must I prepend/append > something "L5" > > when using Intel syntax? (which would seem to > contradict the > > "noprefix" option!) > > > > Luigi > > > > Note: I'm on an intel Mac and both -arch i386 and > -arch x86_64 give > > the same result. _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: GNU assembler freaks out at jumps and Intel syntax
by Terry Lambert 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Ah, you have some reason other than logic for using Intel syntax. Change it to this: 1: L5: popl %ebx cmpl $2, 8(%ebp) je L2 leal LC0-"L00000000001$pb"(%ebx), %eax movl %eax, (%esp) call L_puts$stub .intel_syntax noprefix test %eax, %eax jnz 1b .att_syntax -- Terry This assembles (and links) correctly: testl %eax,%eax jnz L5 This will fail: .intel_syntax noprefix test eax,eax jnz L5 .att_syntax foo.s:28:suffix or operands invalid for `jnz' Luigi _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… On Sep 15, 2008, at 4:15 PM, LuigiG wrote: But that's using the AT&T syntax. That has always worked fine for me too, it's only when you switch to Intel syntax that you get problems. If I chnage your example to use Intel syntax (even limited to that small bit of code), it doesn't work anymore: So, is this a bug in gas? Or must I prepend/append something "L5" when using Intel syntax? (which would seem to contradict the "noprefix" option!) Note: I'm on an intel Mac and both -arch i386 and -arch x86_64 give the same result. This email sent to site_archiver(a)lists.apple.com

1 0

Re: GNU assembler freaks out at jumps and Intel syntax
by LuigiG 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=X-YMail-OSG:Received:X-Mailer:Date:From:Reply-To:Subject:To:MIME-Version:Content-Type:Message-ID; b=eY8Soi5bLEzqXxMhumYy3XlYwNenbrgdj9BOOyGJWpY5gPWObNgNVd4bI9cJkapdnm22+fjTkxnVYFWO409yw+R3e3ZNb4lZafcS5EJYQ1lRsrqYV3ScH0oEVXvG1NyVLvEo0paQfb74fKIlohl0O5BeJsxhmdqSvrr6AtCg8+g=; But that's using the AT&T syntax. That has always worked fine for me too, it's only when you switch to Intel syntax that you get problems. If I chnage your example to use Intel syntax (even limited to that small bit of code), it doesn't work anymore: This assembles (and links) correctly: testl %eax,%eax jnz L5 This will fail: .intel_syntax noprefix test eax,eax jnz L5 .att_syntax foo.s:28:suffix or operands invalid for `jnz' So, is this a bug in gas? Or must I prepend/append something "L5" when using Intel syntax? (which would seem to contradict the "noprefix" option!) Luigi Note: I'm on an intel Mac and both -arch i386 and -arch x86_64 give the same result. --- On Tue, 9/16/08, Terry Lambert <tlambert(a)apple.com> wrote: From: Terry Lambert <tlambert(a)apple.com> Subject: Re: GNU assembler freaks out at jumps and Intel syntax To: news4e71(a)yahoo.com Cc: darwin-dev(a)lists.apple.com Date: Tuesday, September 16, 2008, 12:25 AM On Sep 14, 2008, at 9:04 AM, LuigiG wrote: > Hi, > > What is the correct syntax for conditional jumps (j<cc>) in GNU as > when using intel syntax? > > Specifically: > > # This dummy example code does not assemble: > .intel_syntax noprefix > L1: > test eax,eax > jnz L1 > ret > ..... > %as test1.s > %test1.s:6:suffix or operands invalid for `jnz' > > > # This code (same but temp. switch to AT&T syntax) assembles > correctly: > .intel_syntax noprefix > L1: > test eax,eax > .att_syntax > jnz L1 > .intel_syntax noprefix > ret > > % as test2.s > % otool -Vt test2.o > test2.o: > (__TEXT,__text) section > 00000000 testl %eax,%eax > 00000002 jne 0x00000000 > 00000004 ret > > I'm using: Apple Inc version cctools-698.1~1, GNU assembler version > 1.38 > > What's up?? You are either not specifying -arch i386 and are compiling on a PPC, or you failed to include the .text directive, or you are failing to use cc rather than as to build the .o file from the .s file. Try this with cc -arch i386 -S foo.c: #include <stdio.h> int main(int ac, char *av) { if (ac == 2) goto there; printf("Here\n"); there: printf("There\n"); } Then hack the output .s it to change: leal LC0-"L00000000001$pb"(%ebx), %eax movl %eax, (%esp) call L_puts$stub L2: to: leal LC0-"L00000000001$pb"(%ebx), %eax movl %eax, (%esp) call L_puts$stub testl %eax, %eax jnz L5 L2: And then do cc -arch i386 -o foo foo.s; works fine for me. -- Terry _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: GNU assembler freaks out at jumps and Intel syntax
by LuigiG 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=X-YMail-OSG:Received:X-Mailer:Date:From:Reply-To:Subject:To:MIME-Version:Content-Type:Message-ID; b=rCzwX3e2npkpQfqzf0qF1LFrPEimO1AXBhXqoTb2ly5svsXF8FIY9cALMlFIlnPxU4VrDHKcKkjLazXn5/InZ2VknOQdDY/7wO+QlKB0jcOLy5IJfhfU00iFPK5stgOa2czXkcayevAR0Z+QA2e1TdnnjiB0BwsY7f7OJgOHvAU=; But that's using the AT&T syntax. That has always worked fine for me too, it's only when you switch to Intel syntax that you get problems. If I chnage your example to use Intel syntax (even limited to that small bit of code), it doesn't work anymore: This assembles (and links) correctly: testl %eax,%eax jnz L5 This will fail: .intel_syntax noprefix test eax,eax jnz L5 .att_syntax foo.s:28:suffix or operands invalid for `jnz' So, is this a bug in gas? Or must I prepend/append something "L5" when using Intel syntax? (which would seem to contradict the "noprefix" option!) Luigi Note: I'm on an intel Mac and both -arch i386 and -arch x86_64 give the same result. --- On Tue, 9/16/08, Terry Lambert <tlambert(a)apple.com> wrote: From: Terry Lambert <tlambert(a)apple.com> Subject: Re: GNU assembler freaks out at jumps and Intel syntax To: news4e71(a)yahoo.com Cc: darwin-dev(a)lists.apple.com Date: Tuesday, September 16, 2008, 12:25 AM On Sep 14, 2008, at 9:04 AM, LuigiG wrote: > Hi, > > What is the correct syntax for conditional jumps (j<cc>) in GNU as > when using intel syntax? > > Specifically: > > # This dummy example code does not assemble: > .intel_syntax noprefix > L1: > test eax,eax > jnz L1 > ret > ..... > %as test1.s > %test1.s:6:suffix or operands invalid for `jnz' > > > # This code (same but temp. switch to AT&T syntax) assembles > correctly: > .intel_syntax noprefix > L1: > test eax,eax > .att_syntax > jnz L1 > .intel_syntax noprefix > ret > > % as test2.s > % otool -Vt test2.o > test2.o: > (__TEXT,__text) section > 00000000 testl %eax,%eax > 00000002 jne 0x00000000 > 00000004 ret > > I'm using: Apple Inc version cctools-698.1~1, GNU assembler version > 1.38 > > What's up?? You are either not specifying -arch i386 and are compiling on a PPC, or you failed to include the .text directive, or you are failing to use cc rather than as to build the .o file from the .s file. Try this with cc -arch i386 -S foo.c: #include <stdio.h> int main(int ac, char *av) { if (ac == 2) goto there; printf("Here\n"); there: printf("There\n"); } Then hack the output .s it to change: leal LC0-"L00000000001$pb"(%ebx), %eax movl %eax, (%esp) call L_puts$stub L2: to: leal LC0-"L00000000001$pb"(%ebx), %eax movl %eax, (%esp) call L_puts$stub testl %eax, %eax jnz L5 L2: And then do cc -arch i386 -o foo foo.s; works fine for me. -- Terry _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: GNU assembler freaks out at jumps and Intel syntax
by Terry Lambert 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi, Specifically: # This dummy example code does not assemble: .intel_syntax noprefix L1: test eax,eax jnz L1 ret ..... %as test1.s %test1.s:6:suffix or operands invalid for `jnz' % as test2.s % otool -Vt test2.o test2.o: (__TEXT,__text) section 00000000 testl %eax,%eax 00000002 jne 0x00000000 00000004 ret #include <stdio.h> int main(int ac, char *av) { if (ac == 2) goto there; printf("Here\n"); there: printf("There\n"); } Then hack the output .s it to change: leal LC0-"L00000000001$pb"(%ebx), %eax movl %eax, (%esp) call L_puts$stub L2: to: leal LC0-"L00000000001$pb"(%ebx), %eax movl %eax, (%esp) call L_puts$stub testl %eax, %eax jnz L5 L2: And then do cc -arch i386 -o foo foo.s; works fine for me. -- Terry _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… On Sep 14, 2008, at 9:04 AM, LuigiG wrote: What is the correct syntax for conditional jumps (j<cc>) in GNU as when using intel syntax? # This code (same but temp. switch to AT&T syntax) assembles correctly: .intel_syntax noprefix L1: test eax,eax .att_syntax jnz L1 .intel_syntax noprefix ret I'm using: Apple Inc version cctools-698.1~1, GNU assembler version 1.38 What's up?? You are either not specifying -arch i386 and are compiling on a PPC, or you failed to include the .text directive, or you are failing to use cc rather than as to build the .o file from the .s file. Try this with cc -arch i386 -S foo.c: This email sent to site_archiver(a)lists.apple.com

1 0

Re: launchd wrong behavior for nonexistent paths and crashed services
by Terry Lambert 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hello all, launchd tries to restart indefinitely some service in the following two cases: 1) launchd plist with wrong ProgramArguments key " Throttling respawn: Will start in 10 seconds posix_spawnp("<ProgramArguments>", ...): No such file or directory " The problem is very annoying if some other application waits for this service to start. 2) a service keeps crashing CrashReporter never shows even if I have the CrashReporterPrefs on "Developer" level. It is true that it says: "Display the Unexpectedly Quit dialog for all the processes run by the user". Mac OS X 10.4 had a limit (LAUNCHD_FAILED_EXITS_THRESHOLD). What was the reason to remove it ? -- Terry _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… On Sep 15, 2008, at 9:20 AM, Cristescu Alexandru <cristescualx(a)yahoo.com> wrote: I don't see a reasonable failure mode for this case, since if program B needs service A, it needs service A. You might argue for a cascade failure, where if service A won't load, B should fail in its attempt to contact A rather than launchd answering the phone and being unable to hand off the connection to A. On the other hand, you could argue that while what it's trying to launch doesn't currently exist, it will in the future as a result of a not yet completed operation (a network mount, a disk mount for a disk currently being fsck'ed, a thumb drive containing your ssh keys being inserted, or the binary is in fact being generated, etc.). You could also argue that "the plist needs fixed" and/or "do not install busted plists in the first place". Since there's no way to distinguish a transient vs. Permanent failure, and you presumably intentionally installed it, you intended launchd to keep trying. Is this the system launchd, or the per-session launchd? You will only get the dialog if it's something that's in the current GUI context (search for "CrashReporter" on <http://developer.apple.com> for more details). You could as Dave or file a bug. But see above; there are legitimate reasons for persistently retrying. Probably they outweighed the likelihood of an actually broken plists remaining installed or exporting a service legitimately required by some other process and the system remaining functional. This email sent to site_archiver(a)lists.apple.com

1 0

launchd wrong behavior for nonexistent paths and crashed services
by Cristescu Alexandru 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=X-YMail-OSG:Received:Date:From:Subject:To:MIME-Version:Content-Type:Content-Transfer-Encoding:Message-ID; b=HpvuFpLZPY29zMetCLrNxk6/8JIr1Xr2kSl97gZtulvLDM/WLyQ99aSxxhcgHJ7MRGHnLGaW4aApWbKCEoCli8QNcJTX7lYoyry1lwWlMR6g6vNBGSSrLijvyAo8+9Q06RXcvgwmZd94OYwtHvUU/sZeg4btzgZujqaP4bA8VK8=; Hello all, launchd tries to restart indefinitely some service in the following two cases: 1) launchd plist with wrong ProgramArguments key " Throttling respawn: Will start in 10 seconds posix_spawnp("<ProgramArguments>", ...): No such file or directory " The problem is very annoying if some other application waits for this service to start. 2) a service keeps crashing CrashReporter never shows even if I have the CrashReporterPrefs on "Developer" level. It is true that it says: "Display the Unexpectedly Quit dialog for all the processes run by the user". Mac OS X 10.4 had a limit (LAUNCHD_FAILED_EXITS_THRESHOLD). What was the reason to remove it ? _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Power Management / IOKit
by Godfrey van der Linden 15 Sep '08

15 Sep '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thanks Godfrey On 2008-09-13, at 7:34 PM, Jason Coco wrote: On Sep 13, 2008, at 05:25 , Brian Bergstrand wrote: Jason _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/gvdl%40mac.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… If you have a small test app, can you please raise the bug with bug- reporter. In practice the kernel does usually clean up after a mach- port de-register but you could have detected a corner case. No matter how your app terminates, if you leave dangling resources they will be cleaned up for you by the kernel. There may be a few exceptions, but Mach ports (which is what IOKit handles are) are not one of them. I thought as much, but it seems strange to me that in this case, the computer will no longer go to sleep due to being idle. I've tested it with a simple terminal program. If I interrupt the run loop (^C for example or SIGTERM) and don't call IODeregisterForSystemPower() the computer simply will not sleep again until it's been rebooted. If anyone else could test this, I'd really appreciate it... it seems like a bug but I only have my hardware at the moment to test on and there may be something wrong with that. This email sent to gvdl(a)mac.com This email sent to site_archiver(a)lists.apple.com

1 0