[patch] net/etherboot doesn't compile with gcc4
>Submitter-Id: current-users
>Originator:Jeremie Le Hen
>Organization:
>Confidential: no
>Synopsis: [patch] net/etherboot doesn't compile with gcc4
>Severity: non-critical
>Priority: low
>Category: ports
>Class: sw-bug
>Release: FreeBSD 7.0 i386
>Environment:
System: FreeBSD 7.0
>Description:
GCC 4 is far more nit-picking than its ancestor.
>How-To-Repeat:
Simply try to compile net/etherboot on RELENG_7 or CURRENT.
>Fix:
Drop the following three patchs into the files/ directory.
--- patch-gcc40-1 begins here ---
--- arch/i386/firmware/pcbios/basemem.c.old 2008-02-20 11:24:39.0
+0100
+++ arch/i386/firmware/pcbios/basemem.c 2008-02-20 11:24:49.0 +0100
@@ -93,6 +93,7 @@
uint16_t size_kb = ( size + remainder + 1023 ) >> 10;
free_base_memory_block_t *free_block =
( free_base_memory_block_t * ) ( ptr - remainder );
+ unsigned char *fbaddr;
if ( ( ptr == NULL ) || ( size == 0 ) ) { return; }
@@ -125,7 +126,9 @@
free_block->magic = FREE_BLOCK_MAGIC;
free_block->size_kb = size_kb;
/* Move up by 1 kB */
- (void *)free_block += ( 1 << 10 );
+ fbaddr = (void *)free_block;
+ fbaddr += ( 1 << 10 );
+ free_block = (void *)fbaddr;
size_kb--;
}
--- patch-gcc40-1 ends here ---
--- patch-gcc40-2 begins here ---
--- drivers/net/natsemi.c.old 2008-02-20 11:29:11.0 +0100
+++ drivers/net/natsemi.c 2008-02-20 11:32:13.0 +0100
@@ -602,7 +602,7 @@
const char *p) /* Packet */
{
u32 to, nstype;
-u32 tx_status;
+volatile u32 tx_status;
/* Stop the transmitter */
outl(TxOff, ioaddr + ChipCmd);
@@ -641,7 +641,7 @@
to = currticks() + TX_TIMEOUT;
-while volatile u32) tx_status=txd.cmdsts) & OWN) && (currticks() < to))
+while (((tx_status=txd.cmdsts) & OWN) && (currticks() < to))
/* wait */ ;
if (currticks() >= to) {
--- patch-gcc40-2 ends here ---
--- patch-gcc40-3 begins here ---
--- drivers/net/sis900.c.old2008-02-20 11:33:10.0 +0100
+++ drivers/net/sis900.c2008-02-20 11:33:54.0 +0100
@@ -1083,7 +1083,7 @@
const char *p) /* Packet */
{
u32 to, nstype;
-u32 tx_status;
+volatile u32 tx_status;
/* Stop the transmitter */
outl(TxDIS | inl(ioaddr + cr), ioaddr + cr);
@@ -1122,7 +1122,7 @@
to = currticks() + TX_TIMEOUT;
-while volatile u32) tx_status=txd.cmdsts) & OWN) && (currticks() < to))
+while (((tx_status=txd.cmdsts) & OWN) && (currticks() < to))
/* wait */ ;
if (currticks() >= to) {
--- patch-gcc40-3 ends here ---
___
freebsd-ports@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: Fix for FreeBSD-SA-08:01.pty appears to break net/omnitty?
Rong-En Fan wrote: Hi, On Tue, Feb 19, 2008 at 06:12:59PM -0800, David Wolfskill wrote: On Wed, Feb 20, 2008 at 10:07:07AM +0800, Rong-En Fan wrote: FYI, this is fixed in omnitty 0.3.0_1. Thanks! Should I file a PR against misc/compat6x (because its version of libutil.so is apparently still subject PTY security advisory)? Maintainer is CC'ed. I think we should update misc/compat6x to use contents from 6.3-RELEASE. And filing a PR is also good. Probably i'll update today, maybe tomorrow. Regards Regards, Rong-En Fan -- Marcus Alves Grando marcus(at)sbh.eng.br | Personal mnag(at)FreeBSD.org | FreeBSD.org ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "[EMAIL PROTECTED]"
mail/mail-notification patch 117710 commit?
mail/mail-notification has been broken for a while. The patch attached to [117710][1] worked for me on a fresh ports tree as of 20080220. Perhaps it should be committed/moved into the right queue? [1]: http://www.freebsd.org/cgi/query-pr.cgi?pr=ports/117710 -- <[EMAIL PROTECTED]> "[T]his is not a disentanglement from, but a progressive knotting into." ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "[EMAIL PROTECTED]"
I've started to hate portupgrade
Every new version has such *evident* regressions! Beside needlessly repackaging ports that fail build when doing "-a -p" (obviously non repeatable by the developers), and sometimes failing to reinstall a package if the install phase fails (also non reproducible, but happening nonetheless in production servers). Now, since the last upgrade, I cannot use -R if there's a port with an unknown origin. Look: [EMAIL PROTECTED] portversion -vL= openldap-client-2.3.40 < needs updating (port has 2.3.41) p5-Mail-IMAPClient-2.2.9< [held] needs updating (port has 3.04) p5-PathTools-3.2700 < needs updating (port has 3.2701) p5-Socket6-0.19 < needs updating (port has 0.20) p5-Test-Harness-3.07< needs updating (port has 3.09) squid-2.5.14_4 ? [held] error - origin not found squirrelmail-compatibility-plugin-2.0.9_1 < needs updating (port has 2.0.10) [EMAIL PROTECTED] portupgrade -aRp ** There are errors in a meta info for squid-2.5.14_4 ** Run 'pkgdb -F' to interactively fix them. Of course pkgdb -F asks me to delete squid-2.5.14_4. I also tried to add it to HOLD_PKGS in pkgtools.conf, no change. What the heck does it care if I want to keep an obsolete port! I think the same would happen with a package installed from a developer (non-standard) ports tree, or a perl module installed from CPAN without a corresponding FreeBSD port. Umpf... Any portmaster migration guide for long-time portupgrade users? Angelo. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: I've started to hate portupgrade
Angelo Turetta wrote: Every new version has such *evident* regressions! Beside needlessly repackaging ports that fail build when doing "-a -p" (obviously non repeatable by the developers), and sometimes failing to reinstall a package if the install phase fails (also non reproducible, but happening nonetheless in production servers). Also often happened to me. Portmaster ist my tool of choice, now. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Portupgrade not running get_notinstalled_depends (2.4.3_1, 2 vs 2.4.3_2, 2)
hihi. I was wondering what sort of problems the following change to portupgrade was solving: -- sem 2008-02-11 16:17:39 UTC FreeBSD ports repository Modified files: ports-mgmt/portupgrade Makefile ports-mgmt/portupgrade/files patch-bin-portupgrade Log: - Remove code that gather not installed depends. It makes more problems than solves. Discussed with: pav Revision ChangesPath 1.249 +1 -1 ports/ports-mgmt/portupgrade/Makefile 1.13 +13 -3 ports/ports-mgmt/portupgrade/files/patch-bin-portupgrade -- It seems to break ALT_PKGDEP processing during a portupgrade run because alt_dep happens in get_notinstalled_depends. For example, in my tests, if I have an alt-dep set for a package and I install a port that depends on that package it'll still install the original package rather than my alt-dep. If I revert the change that happened in 2.4.3_2,2 the alt-dep stuff works out. I'm worried that I'm opening up a can of worms because I don't understand what problems were trying to be solved in that version. Any ideas? Or does it sound like I'm misunderstanding ALT_PKGDEP? Thanks! -- James. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "[EMAIL PROTECTED]"
FreeBSD Port: timeseal-1.0 - Bad system call
Please can you help with this? pulsarity# uname -a FreeBSD pulsarity.demax.sk 7.0-RC1 FreeBSD 7.0-RC1 #1: Wed Feb 6 11:45:51 CET 2008 [EMAIL PROTECTED]:/usr/obj/usr/src/sys/PULS i386 pulsarity# timeseal Bad system call pulsarity# kldstat Id Refs AddressSize Name 18 0xc040 60b5b0 kernel 21 0xc0a0c000 30150iwi_bss.ko 31 0xc0a3d000 2f2b0iwi_ibss.ko 41 0xc0a6d000 6b39cacpi.ko 51 0xc7345000 2000 aout.ko This is in 7.0 long time. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: I've started to hate portupgrade
On Wed, 20 Feb 2008 15:25:32 +0100 Angelo Turetta <[EMAIL PROTECTED]> wrote: > Every new version has such *evident* regressions! > > Beside needlessly repackaging ports that fail build when doing "-a > -p" (obviously non repeatable by the developers), and sometimes > failing to reinstall a package if the install phase fails (also non > reproducible, but happening nonetheless in production servers). > > Now, since the last upgrade, I cannot use -R if there's a port with > an unknown origin. > > Look: > > [EMAIL PROTECTED] portversion -vL= > openldap-client-2.3.40 < needs updating (port has 2.3.41) > p5-Mail-IMAPClient-2.2.9< [held] needs updating (port has 3.04) > p5-PathTools-3.2700 < needs updating (port has 3.2701) > p5-Socket6-0.19 < needs updating (port has 0.20) > p5-Test-Harness-3.07< needs updating (port has 3.09) > squid-2.5.14_4 ? [held] error - origin not found > squirrelmail-compatibility-plugin-2.0.9_1 < needs updating (port > has 2.0.10) > > [EMAIL PROTECTED] portupgrade -aRp > ** There are errors in a meta info for squid-2.5.14_4 > ** Run 'pkgdb -F' to interactively fix them. > > Of course pkgdb -F asks me to delete squid-2.5.14_4. I also tried to > add it to HOLD_PKGS in pkgtools.conf, no change. I have had a similar experience with portupgrade from time to time. I made sure I had no reference to it in pkgtools.conf and when I ran pkgdb -F, actually pkgdb -Ffuv, I would choose to delete the entry if I could not get it set correctly. Thereafter, running "portupgrade -apRr" usually worked without any snags. You can always use: "portmanager -u -l -p -bu" to get hard to fix ports working again. At least it has worked for me in the past. Of course, you should run 'pkgdb -Ffu' to insure that portupgrade will operate correctly the next time you invoke it. Anyway, that is my 2¢. -- Gerard [EMAIL PROTECTED] Everybody has something to conceal. Humphrey Bogart signature.asc Description: PGP signature
Re: Portupgrade not running get_notinstalled_depends (2.4.3_1, 2 vs 2.4.3_2, 2)
James píše v st 20. 02. 2008 v 09:32 -0600: > hihi. I was wondering what sort of problems the following change to > portupgrade was solving: > > -- > > sem 2008-02-11 16:17:39 UTC > > FreeBSD ports repository > > Modified files: > ports-mgmt/portupgrade Makefile > ports-mgmt/portupgrade/files patch-bin-portupgrade > Log: > - Remove code that gather not installed depends. It makes more problems > than solves. > > Discussed with: pav > > Revision ChangesPath > 1.249 +1 -1 ports/ports-mgmt/portupgrade/Makefile > 1.13 +13 -3 > ports/ports-mgmt/portupgrade/files/patch-bin-portupgrade > > -- > > It seems to break ALT_PKGDEP processing during a portupgrade run > because alt_dep happens in get_notinstalled_depends. > > For example, in my tests, if I have an alt-dep set for a package and I > install a port that depends on that package it'll still install the > original package rather than my alt-dep. It was installing dependent ports that does not needed to be getting installed. To be more precise, a build dependency of already up-to-date port in a dependency chain of an outdated port. -- Pav Lucistnik <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> See file. Click file. Get file. signature.asc Description: Toto je digitálně podepsaná část zprávy
Re: Portupgrade not running get_notinstalled_depends (2.4.3_1, 2 vs 2.4.3_2, 2)
2008/2/20 Pav Lucistnik <[EMAIL PROTECTED]>: > It was installing dependent ports that does not needed to be getting > installed. To be more precise, a build dependency of already > up-to-date port in a dependency chain of an outdated port. I see. Great. Thanks for the information, Pav. -- James. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to "[EMAIL PROTECTED]"
