Linux 2.4.23-pre1

Linux 2.4.23-pre1

[Marcelo Tosatti]

Hello,

Here goes the first -pre of 2.4.23. It contains a bunch of updates spread
all over, most notably networking.

There are still pending patches in my queue, but I though "Ok, enough
patches for a -pre. "


Detailed changelog below


Summary of changes from v2.4.22 to v2.4.23-pre1
============================================

<achirica:telefonica.net>:
  o [netdrvr airo] Missing defines (only for documentation)
  o [netdrvr airo] MAC type changed to unsigned
  o Missing lines for Wireless Extensions 16
  o MIC support with newer firmware
  o Safer unload code
  o Fix adhoc config

<amir.noam:intel.com>:
  o [net] export alloc_netdev
  o [netdrvr bonding] embed stats struct inside bonding private struct

<davej:redhat.com>:
  o [IPV6]: Missing break in switch statement of rawv6_getsockopt()
  o [IPV4]: /proc/net/pnp dumps items marked initdata

<emann:mrv.com>:
  o [VLAN]: Fix OOPS on module removal

<jan.oravec:6com.sk>:
  o [NET]: Set NLM_F_MULTI in answer of RTM_GETADDR dump answer

<jan:zuchhold.com>:
  o [TG3]: Recognize Altima AC1001 device IDs

<javier:tudela.mad.ttd.net>:
  o [wireless airo] Fixes unregistering of PCI cards
  o [wireless airo] Replaces task queues by simpler kernel_thread

<jdewand:redhat.com>:
  o [SPARC64]: Fix cdrom ioctl32 translations

<kambo77:hotmail.com>:
  o [NET]: Fix hang/memleak in pktgen

<kartik_me:hotmail.com>:
  o [CRYPTO]: Add cast5, integration by jmorris@intercode.com.au

<marcelo:logos.cnet>:
  o pcwd.c: fix oops on unload
  o Cset exclude: m.c.p@wolk-project.de|ChangeSet|20030825183254|28555
  o Cset exclude: m.c.p@wolk-project.de|ChangeSet|20030825194257|34486
  o Fix possible IRQ handling SMP race: Kudos to TeJun Huh
  o Changed EXTRAVERSION to -pre1

<matthewn:snapgear.com>:
  o [netdrvr 8139cp] fix h/w vlan offload

<michel:daenzer.net>:
  o [NET]: Make sure interval member of struct tc_estimator is signed

<mmagallo:debian.org>:
  o AGPGART support for Intel 7x05 chipsets (backported from 2.6)

<skewer:terra.com.br>:
  o [NET]: Remove dead comment from dummy.c driver

<sziwan:hell.org.pl>:
  o [netdrvr 8139too] fix resume behavior, by correctly saving/restoring pci state.

<tv:debian.org>:
  o [NET]: Flush hw header caches on NETDEV_CHANGEADDR events

<wensong:linux-vs.org>:
  o [IPV4]: Add IP Virtual Server to 2.4.x

Alexey Kuznetsov:
  o [IPV4]: IP options were not updated while forwarding multicasts
  o [PKT_SCHED]: More reasonable PSCHED_JSCALE for various values of HZ
  o [IPV4]: Fix rt_score() and usage when purging rtcache hash chains

Andi Kleen:
  o Compile fix for ACPI in 2.4.22/x86-64

Anton Blanchard:
  o [NET]: Add missing memory barriors for __LINK_STATE_RX_SCHED handling

Arjan van de Ven:
  o Fix asm constraint bug in arch/i386/kernel/pci-pc.c

Arnaldo Carvalho de Melo:
  o irqreturn_t compatibility with 2.6

Ben Collins:
  o [SPARC64]: In pci_common.c:find_device_prom_node() recognize PCI_DEVICE_ID_SUN_TOMATILLO
  o [SPARC64]: In clock_probe(), treat m5819p just like m5819

Benjamin Herrenschmidt:
  o [NET]: Do not call request_irq with spinlock held in sungem.c

Chas Williams:
  o [ATM]: export try_atm_clip_ops not atm_clip_ops_mutex

Christoph Hellwig:
  o fix copy_namespace()
  o use list_add_tail in buffer_insert_list
  o reserve a sysctl number for XFS (pagebuf)

Dave Kleikamp:
  o JFS: If unicode conversion fails, operation should fail
  o JFS: Make error return codes negative
  o JFS: K&R to ANSI conversions for fs/jfs/jfs_dmap.c and jfs_xtree.c
  o JFS: add nointegrity mount option (Karl Rister)

David S. Miller:
  o [SPARC64]: Add Ultra-IIIi/Jalapeno support
  o [SPARC64]: Add JIO/Tomatillo PCI controller support
  o [SPARC64]: Read processor number correctly on Ultra-IIIi/Jalapeno
  o [SPARC64]: In ISA support, is interrupt-map exists use it
  o [SPARC64]: Finalize TOMATILLO/JIO support, help from bcollins@debian.org
  o [TG3]: Support OBP firmware mac-addresses on sparc64
  o [SPARC64]: Sanitize PCI controller handling to support Tomatillo better
  o [SPARC64]: Pass correct args to data_access_exception() in unaligned.c
  o [SPARC64]: Make sure to reject all PCI DAC dma masks
  o [SPARC64]: In schizo driver, if virtual-dma property exists, respect it
  o [ATM]: Remove -g option from driver directory CFLAGS
  o [SPARC64]: More tomatillo PCI controller fixes
  o [TG3]: More Sun onboard 5704 fixes
  o [TG3]: Only call tg3_init_rings() after hardware has been reset
  o [SPARC64]: Fix AFSR error reporting for Cheetah+/Jalapeno
  o [SPARC64]: Missing cheetah+ ASI defines
  o [SPARC64]: Fix unused variable warnings when using iounmap()
  o [SPARC64]: Do not make sparc_{cpu,fpu}_type a NR_CPUS array
  o [NET]: Export neigh_changeaddr
  o [SPARC64]: Add some missing PCI error reporting
  o [TG3]: Fix AC1001 typo in pci_ids.h
  o [NET]: Include asm/uaccess.h in net/core/ethtool.c

Harald Welte:
  o [NETFILTER]: Backport iptables AH/ESP fixes from 2.6.x
  o [NETFILTER]: Fix uninitialized return in iptables tftp
  o [NETFILTER]: NAT optimization
  o [NETFILTER]: Conntrack optimization (LIST_DELETE)

Hideaki Yoshifuji:
  o [IPV6]: Fix typo in linux/ipv6.h

Ion Badulescu:
  o [netdrvr tulip] add pci id for 3com 3CSOHO100B-TX

Jack Hammer:
  o ServeRAID 6.10 Driver Update

Jeff Garzik:
  o [ia32] Via, Intel cpu capabilities update
  o [hw_random] add combined Intel+AMD+VIA h/w RNG driver
  o [NET]: Backport ethtool_ops from 2.6.x
  o [ia32] mention that X86_VENDOR_ID is tied to NCAPINTS, in a comment in arch/i386/kernel/head.S.

John Stultz:
  o Do not clear SMI pin at bootup
  o Handle clustered XAPIC in set_ioapic_affinity()

Keith M. Wesolowski:
  o [SPARC]: Trap table alignment for HyperSPARC

Krishna Kumar:
  o [IPV6]: Reporting of prefix routes via rtnetlink

Marc-Christian Petersen:
  o Cleanup kmem_cache_reap()
  o Fix initrd with netboot
  o Cleanup DRM submenu
  o Replace bogus and obsolete "#if __SMP__" -> CONFIG_SMP
  o Allow console switching after kernel panic()
  o Unblank console if panic()
  o Handle get_block errors correctly in block_read_full_page()
  o LVM Update v1.0.5 to v1.0.7
  o CONFIG_NR_CPUS
  o Avoid potentially leaking pagetables into the per-cpu queues
  o Proper APIC with HyperThreading

Martin Devera:
  o [NET]: Fix bugs in sch_htb packet scheduler

Mikael Pettersson:
  o [ia32] adjust X86_VENDOR_ID offset in head.S, due to new NCAPINTS
  o 2.4.22 local APIC updates 1/3: remove incorrect blacklist rules
  o 2.4.22 local APIC updates 2/3: add lapic/nolapic options
  o 2.4.22 local APIC updates 3/3: disable APIC_BASE on reboot

Patrick McHardy:
  o [NET]: Fix no_cong_thresh sysctl

Randy Dunlap:
  o [NET]: Audit copy_from_user checks in pktgen

Robert Olsson:
  o [NET]: Remove some debugging from pktgen

Rusty Russell:
  o [NETFILTER]: Fix masquerade routing check, backport to 2.4 by kurd@cp.rtfm.se

Stelian Pop:
  o sonypi driver update
  o meye driver updates

Stephen Hemminger:
  o [BRIDGE]: Mailing list is at osdl.org now
  o [VLAN]: Allow it to compile with VLAN_DEBUG enabled

Willy Tarreau:
  o Fix amd67x_pm.c crash with no chipsets / CONFIG_HOTPLUG
  o make log buffer length selectable

Re: Linux 2.4.23-pre1

[Willy Tarreau]

Hi Marcelo,

On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
> Willy Tarreau:
>   o Fix amd67x_pm.c crash with no chipsets / CONFIG_HOTPLUG

Was mostly from Herbert Xu.

Cheers,
Willy

Spam? was: Linux 2.4.23-pre1

[Mike Fedyk]

On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
> This mail is probably spam.  The original message has been attached
> along with this report, so you can recognize or block similar unwanted
> mail in future.  See http://spamassassin.org/tag/ for more details.
> 
> Content preview:  Hello, Here goes the first -pre of 2.4.23. It contains
>   a bunch of updates spread all over, most notably networking. There are
>   still pending patches in my queue, but I though "Ok, enough patches for
>   a -pre. " [...] 
> 
> Content analysis details:   (6.00 points, 5 required)
> USER_AGENT_PINE    (0.0 points)  Message-Id indicates a non-spam MUA (Pine)
> X_MAILING_LIST     (0.0 points)  Has a X-Mailing-List header
> ROUND_THE_WORLD    (2.5 points)  Received: says mail bounced around the world (DNS)
> RCVD_IN_OSIRUSOFT_COM (0.6 points)  RBL: Received via a relay in relays.osirusoft.com
>                    [RBL check: found 212.78.72.67.relays.osirusoft.com.]
> X_OSIRU_OPEN_RELAY (2.9 points)  RBL: DNSBL: sender is Confirmed Open Relay

This was marked as spam from spamassassin.

Note the entry in RBL for this server...

Just letting you guys know.

Mike

Re: Spam? was: Linux 2.4.23-pre1

[Richard B. Johnson]

On Thu, 28 Aug 2003, Mike Fedyk wrote:

> On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
> > This mail is probably spam.  The original message has been attached
> > along with this report, so you can recognize or block similar unwanted
> > mail in future.  See http://spamassassin.org/tag/ for more details.
> >
> > Content preview:  Hello, Here goes the first -pre of 2.4.23. It contains
> >   a bunch of updates spread all over, most notably networking. There are
> >   still pending patches in my queue, but I though "Ok, enough patches for
> >   a -pre. " [...]
> >
> > Content analysis details:   (6.00 points, 5 required)
> > USER_AGENT_PINE    (0.0 points)  Message-Id indicates a non-spam MUA (Pine)
> > X_MAILING_LIST     (0.0 points)  Has a X-Mailing-List header
> > ROUND_THE_WORLD    (2.5 points)  Received: says mail bounced around the world (DNS)
> > RCVD_IN_OSIRUSOFT_COM (0.6 points)  RBL: Received via a relay in relays.osirusoft.com
> >                    [RBL check: found 212.78.72.67.relays.osirusoft.com.]
> > X_OSIRU_OPEN_RELAY (2.9 points)  RBL: DNSBL: sender is Confirmed Open Relay
>
> This was marked as spam from spamassassin.
>
> Note the entry in RBL for this server...
>
> Just letting you guys know.
>
> Mike

Lucky you didn't send the source code aic7xxx.c! This is
known porographic material!

Cheers,
Dick Johnson
Penguin : Linux version 2.4.22 on an i686 machine (794.73 BogoMips).
            Note 96.31% of all statistics are fiction.

Re: Spam? was: Linux 2.4.23-pre1

[Alan Cox]

On Iau, 2003-08-28 at 18:42, Mike Fedyk wrote:
> On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
> > This mail is probably spam.  The original message has been attached
> > along with this report, so you can recognize or block similar unwanted

Your bug, osirisoft reports all mail as spam now and all hosts as
spam sources. Its probably the most accurate reporting its ever done 8)

Re: Spam? was: Linux 2.4.23-pre1

[John Jasen]

On Thu, 28 Aug 2003, Mike Fedyk wrote:
> On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
> > Content analysis details:   (6.00 points, 5 required)
> > USER_AGENT_PINE    (0.0 points)  Message-Id indicates a non-spam MUA (Pine)
> > X_MAILING_LIST     (0.0 points)  Has a X-Mailing-List header
> > ROUND_THE_WORLD    (2.5 points)  Received: says mail bounced around the world (DNS)
> > RCVD_IN_OSIRUSOFT_COM (0.6 points)  RBL: Received via a relay in relays.osirusoft.com
> >                    [RBL check: found 212.78.72.67.relays.osirusoft.com.]
> > X_OSIRU_OPEN_RELAY (2.9 points)  RBL: DNSBL: sender is Confirmed Open Relay
> 
> This was marked as spam from spamassassin.
> 
> Note the entry in RBL for this server...

Osirusoft RBL is effectively dead, and should no longer be used for 
filtering or tagging -- last I heard, everyone was getting listed.

-- 
-- John E. Jasen (jjasen@realityfailure.org)
-- User Error #2361: Please insert coffee and try again.

Re: Spam? was: Linux 2.4.23-pre1

[Stan Bubrouski]

John Jasen wrote:
> On Thu, 28 Aug 2003, Mike Fedyk wrote:
> 
>>On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
>>
>>>Content analysis details:   (6.00 points, 5 required)
>>>USER_AGENT_PINE    (0.0 points)  Message-Id indicates a non-spam MUA (Pine)
>>>X_MAILING_LIST     (0.0 points)  Has a X-Mailing-List header
>>>ROUND_THE_WORLD    (2.5 points)  Received: says mail bounced around the world (DNS)
>>>RCVD_IN_OSIRUSOFT_COM (0.6 points)  RBL: Received via a relay in relays.osirusoft.com
>>>                   [RBL check: found 212.78.72.67.relays.osirusoft.com.]
>>>X_OSIRU_OPEN_RELAY (2.9 points)  RBL: DNSBL: sender is Confirmed Open Relay
>>
>>This was marked as spam from spamassassin.
>>
>>Note the entry in RBL for this server...
> 
> 
> Osirusoft RBL is effectively dead, and should no longer be used for 
> filtering or tagging -- last I heard, everyone was getting listed.
> 

As previously stated in a different thread Osirusoft is
dead and currently blocking *.*.*.* to indicate this
position.  On the other hand, if you use it, you won't
need any other filering ;-)

-sb

Re: Spam? was: Linux 2.4.23-pre1

[Adam Sampson]

Mike Fedyk <mfedyk@matchmail.com> writes:

>> RCVD_IN_OSIRUSOFT_COM (0.6 points) RBL: Received via a relay in
>> relays.osirusoft.com
> This was marked as spam from spamassassin.

relays.osirusoft.com have closed down, and are now apparently
reporting all addresses as blacklisted (which strikes me as not
exactly the most user-friendly way to shut down a blacklist, but I
guess they've got their reasons...).

-- 
Adam Sampson <azz@us-lot.org>                        <http://offog.org/>

Re: Spam? was: Linux 2.4.23-pre1

[Marc Schiffbauer]

* Mike Fedyk schrieb am 28.08.03 um 19:42 Uhr:
> On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
> > This mail is probably spam.  The original message has been attached
> > along with this report, so you can recognize or block similar unwanted
> > mail in future.  See http://spamassassin.org/tag/ for more details.
> > 
> > Content preview:  Hello, Here goes the first -pre of 2.4.23. It contains
> >   a bunch of updates spread all over, most notably networking. There are
> >   still pending patches in my queue, but I though "Ok, enough patches for
> >   a -pre. " [...] 
> > 
> > Content analysis details:   (6.00 points, 5 required)
> > USER_AGENT_PINE    (0.0 points)  Message-Id indicates a non-spam MUA (Pine)
> > X_MAILING_LIST     (0.0 points)  Has a X-Mailing-List header
> > ROUND_THE_WORLD    (2.5 points)  Received: says mail bounced around the world (DNS)
> > RCVD_IN_OSIRUSOFT_COM (0.6 points)  RBL: Received via a relay in relays.osirusoft.com
> >                    [RBL check: found 212.78.72.67.relays.osirusoft.com.]
> > X_OSIRU_OPEN_RELAY (2.9 points)  RBL: DNSBL: sender is Confirmed Open Relay
> 
> This was marked as spam from spamassassin.
> 
> Note the entry in RBL for this server...
> 
> Just letting you guys know.
> 

This is because the osirusoft.com RBL server has been under attack
(ddos) last weekend. Answers by this server were always "listed"
during this time. This is not a lucky default IMHO but it was like
that.

See also: http://www.heise.de/newsticker/data/uma-27.08.03-000/
(for those who read german)

For me it was worse because my exim used this server directly and
did not mark those mails as spam but completely blocked all mails
:-(

BTW: Does newest release contain this rules (*_OSIRUSOFT_*) by
default?

-Marc

-- 
8AAC 5F46 83B4 DB70 8317  3723 296C 6CCA 35A6 4134

Re: Spam? was: Linux 2.4.23-pre1

[Mike Fedyk]

On Thu, Aug 28, 2003 at 06:57:01PM +0100, Alan Cox wrote:
> On Iau, 2003-08-28 at 18:42, Mike Fedyk wrote:
> > On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
> > > This mail is probably spam.  The original message has been attached
> > > along with this report, so you can recognize or block similar unwanted
> 
> Your bug, osirisoft reports all mail as spam now and all hosts as
> spam sources. Its probably the most accurate reporting its ever done 8)

Yep, very true.

Just checked the debian bug track system, and there is an upload that fixes
this today.

Thanks

DO NOT USE osirusoft.com DATASETS!

[Matti Aarnio]

On Thu, Aug 28, 2003 at 10:42:47AM -0700, Mike Fedyk wrote:
> > Content analysis details:   (6.00 points, 5 required)
> > USER_AGENT_PINE    (0.0 points)  Message-Id indicates a non-spam MUA (Pine)
> > X_MAILING_LIST     (0.0 points)  Has a X-Mailing-List header
> > ROUND_THE_WORLD    (2.5 points)  Received: says mail bounced around world (DNS)
> > RCVD_IN_OSIRUSOFT_COM (0.6 points)  RBL: Received via a relay in relays.osirusoft.com
> >                    [RBL check: found 212.78.72.67.relays.osirusoft.com.]
> > X_OSIRU_OPEN_RELAY (2.9 points)  RBL: DNSBL: sender is Confirmed Open Relay

  So you are using OSIRUSOFT's databases ?
  Haven't noticed that they have been poisoned/broken for past few days ?

> Note the entry in RBL for this server...
> Just letting you guys know.
> Mike

[2.4 patch] Fix IRQ_NONE clash in SCSI drivers

[Adrian Bunk]

On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
>...
> Summary of changes from v2.4.22 to v2.4.23-pre1
> ============================================
>...
> Arnaldo Carvalho de Melo:
>   o irqreturn_t compatibility with 2.6
>...

This change added an (empty) IRQ_NONE #define to interrupt.h.

Several scsi drivers are already using an IRQ_NONE.  Rename that to
SCSI_IRQ_NONE (a similar change was done in 2.5 by Andrew Morton several
months ago).

I've tested the compilation with 2.4.23-pre2.

Please apply
Adrian

--- linux-2.4.23-pre2-full/drivers/scsi/53c7,8xx.c.old	2003-09-02 20:27:14.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/53c7,8xx.c	2003-09-02 20:27:25.000000000 +0200
@@ -6397,7 +6397,7 @@
 	(struct NCR53c7x0_hostdata *) host->hostdata;
     struct NCR53c7x0_cmd *cmd, *tmp;
     shutdown (host);
-    if (host->irq != IRQ_NONE)
+    if (host->irq != SCSI_IRQ_NONE)
 	{
 	    int irq_count;
 	    struct Scsi_Host *tmp;
--- linux-2.4.23-pre2-full/drivers/scsi/53c7,8xx.h.old	2003-09-02 20:27:25.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/53c7,8xx.h	2003-09-02 20:27:33.000000000 +0200
@@ -1428,7 +1428,7 @@
 
 };
 
-#define IRQ_NONE	255
+#define SCSI_IRQ_NONE	255
 #define DMA_NONE	255
 #define IRQ_AUTO	254
 #define DMA_AUTO	254
--- linux-2.4.23-pre2-full/drivers/scsi/53c7xx.c.old	2003-09-02 20:27:33.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/53c7xx.c	2003-09-02 20:27:42.000000000 +0200
@@ -6081,7 +6081,7 @@
 	(struct NCR53c7x0_hostdata *) host->hostdata[0];
     struct NCR53c7x0_cmd *cmd, *tmp;
     shutdown (host);
-    if (host->irq != IRQ_NONE)
+    if (host->irq != SCSI_IRQ_NONE)
 	{
 	    int irq_count;
 	    struct Scsi_Host *tmp;
--- linux-2.4.23-pre2-full/drivers/scsi/53c7xx.h.old	2003-09-02 20:27:42.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/53c7xx.h	2003-09-02 20:27:49.000000000 +0200
@@ -1450,7 +1450,7 @@
 
 };
 
-#define IRQ_NONE	255
+#define SCSI_IRQ_NONE	255
 #define DMA_NONE	255
 #define IRQ_AUTO	254
 #define DMA_AUTO	254
--- linux-2.4.23-pre2-full/drivers/scsi/NCR5380.c.old	2003-09-02 20:27:49.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/NCR5380.c	2003-09-02 20:32:27.000000000 +0200
@@ -693,7 +693,7 @@
 			trying_irqs |= mask;
 
 	timeout = jiffies + (250 * HZ / 1000);
-	probe_irq = IRQ_NONE;
+	probe_irq = SCSI_IRQ_NONE;
 
 	/*
 	 * A interrupt is triggered whenever BSY = false, SEL = true
@@ -710,7 +710,7 @@
 	NCR5380_write(OUTPUT_DATA_REG, hostdata->id_mask);
 	NCR5380_write(INITIATOR_COMMAND_REG, ICR_BASE | ICR_ASSERT_DATA | ICR_ASSERT_SEL);
 
-	while (probe_irq == IRQ_NONE && time_before(jiffies, timeout))
+	while (probe_irq == SCSI_IRQ_NONE && time_before(jiffies, timeout))
 		barrier();
 
 	NCR5380_write(SELECT_ENABLE_REG, 0);
@@ -860,7 +860,7 @@
 
 	SPRINTF("\nBase Addr: 0x%05lX    ", (long) instance->base);
 	SPRINTF("io_port: %04x      ", (int) instance->io_port);
-	if (instance->irq == IRQ_NONE)
+	if (instance->irq == SCSI_IRQ_NONE)
 		SPRINTF("IRQ: None.\n");
 	else
 		SPRINTF("IRQ: %d.\n", instance->irq);
@@ -1706,7 +1706,7 @@
 #endif				/* def NCR_TIMEOUT */
 
 	dprintk(NDEBUG_SELECTION, ("scsi%d : target %d selected, going into MESSAGE OUT phase.\n", instance->host_no, cmd->target));
-	tmp[0] = IDENTIFY(((instance->irq == IRQ_NONE) ? 0 : 1), cmd->lun);
+	tmp[0] = IDENTIFY(((instance->irq == SCSI_IRQ_NONE) ? 0 : 1), cmd->lun);
 
 	len = 1;
 	cmd->tag = 0;
--- linux-2.4.23-pre2-full/drivers/scsi/dmx3191d.c.old	2003-09-02 20:27:59.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/dmx3191d.c	2003-09-02 20:33:02.000000000 +0200
@@ -96,7 +96,7 @@
 			printk(KERN_WARNING "dmx3191: IRQ %d not available - switching to polled mode.\n", pdev->irq);
 			/* Steam powered scsi controllers run without an IRQ
 			   anyway */
-			instance->irq = IRQ_NONE;
+			instance->irq = SCSI_IRQ_NONE;
 		}
 
 		boards++;
@@ -113,7 +113,7 @@
 int dmx3191d_release_resources(struct Scsi_Host *instance)
 {
 	release_region(instance->io_port, DMX3191D_REGION);
-	if(instance->irq!=IRQ_NONE)
+	if(instance->irq != SCSI_IRQ_NONE)
 		free_irq(instance->irq, instance);
 
 	return 0;
--- linux-2.4.23-pre2-full/drivers/scsi/dtc.c.old	2003-09-02 20:28:12.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/dtc.c	2003-09-02 20:29:03.000000000 +0200
@@ -261,24 +261,24 @@
 #ifndef DONT_USE_INTR
 		/* With interrupts enabled, it will sometimes hang when doing heavy
 		 * reads. So better not enable them until I figure it out. */
-		if (instance->irq != IRQ_NONE)
+		if (instance->irq != SCSI_IRQ_NONE)
 			if (request_irq(instance->irq, do_dtc_intr, SA_INTERRUPT, "dtc")) 
 			{
 				printk(KERN_WARNING "scsi%d : IRQ%d not free, interrupts disabled\n", instance->host_no, instance->irq);
-				instance->irq = IRQ_NONE;
+				instance->irq = SCSI_IRQ_NONE;
 			}
 
-		if (instance->irq == IRQ_NONE) {
+		if (instance->irq == SCSI_IRQ_NONE) {
 			printk(KERN_INFO "scsi%d : interrupts not enabled. for better interactive performance,\n", instance->host_no);
 			printk(KERN_INFO "scsi%d : please jumper the board for a free IRQ.\n", instance->host_no);
 		}
 #else
-		if (instance->irq != IRQ_NONE)
+		if (instance->irq != SCSI_IRQ_NONE)
 			printk(KERN_INFO "scsi%d : interrupts not used. Might as well not jumper it.\n", instance->host_no);
-		instance->irq = IRQ_NONE;
+		instance->irq = SCSI_IRQ_NONE;
 #endif
 		printk(KERN_INFO "scsi%d : at 0x%05X", instance->host_no, (int) instance->base);
-		if (instance->irq == IRQ_NONE)
+		if (instance->irq == SCSI_IRQ_NONE)
 			printk(" interrupts disabled");
 		else
 			printk(" irq %d", instance->irq);
@@ -336,7 +336,7 @@
 	i = 0;
 	NCR5380_read(RESET_PARITY_INTERRUPT_REG);
 	NCR5380_write(MODE_REG, MR_ENABLE_EOP_INTR | MR_DMA_MODE);
-	if (instance->irq == IRQ_NONE)
+	if (instance->irq == SCSI_IRQ_NONE)
 		NCR5380_write(DTC_CONTROL_REG, CSR_DIR_READ);
 	else
 		NCR5380_write(DTC_CONTROL_REG, CSR_DIR_READ | CSR_INT_BASE);
@@ -384,7 +384,7 @@
 	NCR5380_read(RESET_PARITY_INTERRUPT_REG);
 	NCR5380_write(MODE_REG, MR_ENABLE_EOP_INTR | MR_DMA_MODE);
 	/* set direction (write) */
-	if (instance->irq == IRQ_NONE)
+	if (instance->irq == SCSI_IRQ_NONE)
 		NCR5380_write(DTC_CONTROL_REG, 0);
 	else
 		NCR5380_write(DTC_CONTROL_REG, CSR_5380_INTR);
--- linux-2.4.23-pre2-full/drivers/scsi/g_NCR5380.c.old	2003-09-02 20:29:03.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/g_NCR5380.c	2003-09-02 20:33:26.000000000 +0200
@@ -333,7 +333,7 @@
 			if (dev->irq_resource[0].flags & IORESOURCE_IRQ)
 				overrides[count].irq = dev->irq_resource[0].start;
 			else
-				overrides[count].irq = IRQ_NONE;
+				overrides[count].irq = SCSI_IRQ_NONE;
 			if (dev->dma_resource[0].flags & IORESOURCE_DMA)
 				overrides[count].dma = dev->dma_resource[0].start;
 			else
@@ -430,19 +430,19 @@
 		else
 			instance->irq = NCR5380_probe_irq(instance, 0xffff);
 
-		if (instance->irq != IRQ_NONE)
+		if (instance->irq != SCSI_IRQ_NONE)
 			if (request_irq(instance->irq, do_generic_NCR5380_intr, SA_INTERRUPT, "NCR5380", NULL)) {
 				printk(KERN_WARNING "scsi%d : IRQ%d not free, interrupts disabled\n", instance->host_no, instance->irq);
-				instance->irq = IRQ_NONE;
+				instance->irq = SCSI_IRQ_NONE;
 			}
 
-		if (instance->irq == IRQ_NONE) {
+		if (instance->irq == SCSI_IRQ_NONE) {
 			printk(KERN_INFO "scsi%d : interrupts not enabled. for better interactive performance,\n", instance->host_no);
 			printk(KERN_INFO "scsi%d : please jumper the board for a free IRQ.\n", instance->host_no);
 		}
 
 		printk(KERN_INFO "scsi%d : at " STRVAL(NCR5380_map_name) " 0x%x", instance->host_no, (unsigned int) instance->NCR5380_instance_name);
-		if (instance->irq == IRQ_NONE)
+		if (instance->irq == SCSI_IRQ_NONE)
 			printk(" interrupts disabled");
 		else
 			printk(" irq %d", instance->irq);
@@ -489,7 +489,7 @@
 	release_mem_region(instance->NCR5380_instance_name, NCR5380_region_size);
 #endif
 
-	if (instance->irq != IRQ_NONE)
+	if (instance->irq != SCSI_IRQ_NONE)
 		free_irq(instance->irq, NULL);
 
 	return 0;
@@ -802,7 +802,7 @@
 	PRINTP("NO NCR53C400 driver extensions\n");
 #endif
 	PRINTP("Using %s mapping at %s 0x%lx, " ANDP STRVAL(NCR5380_map_config) ANDP STRVAL(NCR5380_map_name) ANDP scsi_ptr->NCR5380_instance_name);
-	if (scsi_ptr->irq == IRQ_NONE)
+	if (scsi_ptr->irq == SCSI_IRQ_NONE)
 		PRINTP("no interrupt\n");
 	else
 		PRINTP("on interrupt %d\n" ANDP scsi_ptr->irq);
--- linux-2.4.23-pre2-full/drivers/scsi/mac_scsi.c.old	2003-09-02 20:29:11.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/mac_scsi.c	2003-09-02 20:29:32.000000000 +0200
@@ -253,16 +253,16 @@
 
     ((struct NCR5380_hostdata *)instance->hostdata)->ctrl = 0;
 
-    if (instance->irq != IRQ_NONE)
+    if (instance->irq != SCSI_IRQ_NONE)
 	if (request_irq(instance->irq, NCR5380_intr, IRQ_FLG_SLOW, 
 		"ncr5380", NCR5380_intr)) {
 	    printk(KERN_WARNING "scsi%d: IRQ%d not free, interrupts disabled\n",
 		   instance->host_no, instance->irq);
-	    instance->irq = IRQ_NONE;
+	    instance->irq = SCSI_IRQ_NONE;
 	}
 
     printk(KERN_INFO "scsi%d: generic 5380 at port %lX irq", instance->host_no, instance->io_port);
-    if (instance->irq == IRQ_NONE)
+    if (instance->irq == SCSI_IRQ_NONE)
 	printk (KERN_INFO "s disabled");
     else
 	printk (KERN_INFO " %d", instance->irq);
@@ -277,7 +277,7 @@
 
 int macscsi_release (struct Scsi_Host *shpnt)
 {
-	if (shpnt->irq != IRQ_NONE)
+	if (shpnt->irq != SCSI_IRQ_NONE)
 		free_irq (shpnt->irq, NCR5380_intr);
 
 	return 0;
--- linux-2.4.23-pre2-full/drivers/scsi/pas16.c.old	2003-09-02 20:29:32.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/pas16.c	2003-09-02 20:30:00.000000000 +0200
@@ -99,11 +99,11 @@
  *   interrupts.  Ie, for a board at the default 0x388 base port,
  *   boot: linux pas16=0x388,255
  *
- *   IRQ_NONE (255) should be specified for no interrupt,
+ *   SCSI_IRQ_NONE (255) should be specified for no interrupt,
  *   IRQ_AUTO (254) to autoprobe for an IRQ line if overridden
  *   on the command line.
  *
- *   (IRQ_AUTO == 254, IRQ_NONE == 255 in NCR5380.h)
+ *   (IRQ_AUTO == 254, SCSI_IRQ_NONE == 255 in NCR5380.h)
  */
 
 #include <linux/module.h>
@@ -404,13 +404,13 @@
 		else
 			instance->irq = NCR5380_probe_irq(instance, PAS16_IRQS);
 
-		if (instance->irq != IRQ_NONE)
+		if (instance->irq != SCSI_IRQ_NONE)
 			if (request_irq(instance->irq, do_pas16_intr, SA_INTERRUPT, "pas16", NULL)) {
 				printk(KERN_WARNING "scsi%d : IRQ%d not free, interrupts disabled\n", instance->host_no, instance->irq);
-				instance->irq = IRQ_NONE;
+				instance->irq = SCSI_IRQ_NONE;
 			}
 
-		if (instance->irq == IRQ_NONE) {
+		if (instance->irq == SCSI_IRQ_NONE) {
 			printk(KERN_INFO "scsi%d : interrupts not enabled. for better interactive performance,\n", instance->host_no);
 			printk(KERN_INFO "scsi%d : please jumper the board for a free IRQ.\n", instance->host_no);
 			/* Disable 5380 interrupts, leave drive params the same */
@@ -420,7 +420,7 @@
 
 		printk(KERN_INFO "scsi%d : at 0x%04x", instance->host_no, (int)
 		       instance->io_port);
-		if (instance->irq == IRQ_NONE)
+		if (instance->irq == SCSI_IRQ_NONE)
 			printk(" interrupts disabled");
 		else
 			printk(" irq %d", instance->irq);
--- linux-2.4.23-pre2-full/drivers/scsi/sun3_scsi.c.old	2003-09-02 20:30:00.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/sun3_scsi.c	2003-09-02 20:30:16.000000000 +0200
@@ -267,7 +267,7 @@
 #ifndef REAL_DMA
 		printk("scsi%d: IRQ%d not free, interrupts disabled\n",
 		       instance->host_no, instance->irq);
-		instance->irq = IRQ_NONE;
+		instance->irq = SCSI_IRQ_NONE;
 #else
 		printk("scsi%d: IRQ%d not free, bailing out\n",
 		       instance->host_no, instance->irq);
@@ -276,7 +276,7 @@
 	}
 	
 	printk("scsi%d: Sun3 5380 at port %lX irq", instance->host_no, instance->io_port);
-	if (instance->irq == IRQ_NONE)
+	if (instance->irq == SCSI_IRQ_NONE)
 		printk ("s disabled");
 	else
 		printk (" %d", instance->irq);
@@ -305,7 +305,7 @@
 #ifdef MODULE
 int sun3scsi_release (struct Scsi_Host *shpnt)
 {
-	if (shpnt->irq != IRQ_NONE)
+	if (shpnt->irq != SCSI_IRQ_NONE)
 		free_irq (shpnt->irq, NULL);
 
 	iounmap((void *)sun3_scsi_regp);
--- linux-2.4.23-pre2-full/drivers/scsi/sun3_scsi_vme.c.old	2003-09-02 20:30:16.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/sun3_scsi_vme.c	2003-09-02 20:30:37.000000000 +0200
@@ -236,7 +236,7 @@
 #ifndef REAL_DMA
 		printk("scsi%d: IRQ%d not free, interrupts disabled\n",
 		       instance->host_no, instance->irq);
-		instance->irq = IRQ_NONE;
+		instance->irq = SCSI_IRQ_NONE;
 #else
 		printk("scsi%d: IRQ%d not free, bailing out\n",
 		       instance->host_no, instance->irq);
@@ -245,7 +245,7 @@
 	}
 
 	printk("scsi%d: Sun3 5380 VME at port %lX irq", instance->host_no, instance->io_port);
-	if (instance->irq == IRQ_NONE)
+	if (instance->irq == SCSI_IRQ_NONE)
 		printk ("s disabled");
 	else
 		printk (" %d", instance->irq);
@@ -281,7 +281,7 @@
 #ifdef MODULE
 int sun3scsi_release (struct Scsi_Host *shpnt)
 {
-	if (shpnt->irq != IRQ_NONE)
+	if (shpnt->irq != SCSI_IRQ_NONE)
 		free_irq (shpnt->irq, NULL);
 
 	iounmap(sun3_scsi_regp);
--- linux-2.4.23-pre2-full/drivers/scsi/t128.c.old	2003-09-02 20:30:37.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/t128.c	2003-09-02 20:30:59.000000000 +0200
@@ -237,20 +237,20 @@
 		else
 			instance->irq = NCR5380_probe_irq(instance, T128_IRQS);
 
-		if (instance->irq != IRQ_NONE)
+		if (instance->irq != SCSI_IRQ_NONE)
 			if (request_irq(instance->irq, do_t128_intr, SA_INTERRUPT, "t128", NULL)) 
 			{
 				printk(KERN_WARNING "scsi%d : IRQ%d not free, interrupts disabled\n", instance->host_no, instance->irq);
-				instance->irq = IRQ_NONE;
+				instance->irq = SCSI_IRQ_NONE;
 			}
 
-		if (instance->irq == IRQ_NONE) {
+		if (instance->irq == SCSI_IRQ_NONE) {
 			printk(KERN_INFO "scsi%d : interrupts not enabled. for better interactive performance,\n", instance->host_no);
 			printk(KERN_INFO "scsi%d : please jumper the board for a free IRQ.\n", instance->host_no);
 		}
 
 		printk(KERN_INFO "scsi%d : at 0x%08lx", instance->host_no,instance->base);
-		if (instance->irq == IRQ_NONE)
+		if (instance->irq == SCSI_IRQ_NONE)
 			printk(" interrupts disabled");
 		else
 			printk(" irq %d", instance->irq);
--- linux-2.4.23-pre2-full/drivers/scsi/tmscsim.c.old	2003-09-02 20:30:59.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/tmscsim.c	2003-09-02 20:31:05.000000000 +0200
@@ -3047,7 +3047,7 @@
     /* TO DO: We should check for outstanding commands first. */
     dc390_shutdown (host);
 
-    if (host->irq != IRQ_NONE)
+    if (host->irq != SCSI_IRQ_NONE)
     {
 	DEBUG0(printk(KERN_INFO "DC390: Free IRQ %i\n",host->irq);)
 	free_irq (host->irq, pACB);
--- linux-2.4.23-pre2-full/drivers/scsi/tmscsim.h.old	2003-09-02 20:31:05.000000000 +0200
+++ linux-2.4.23-pre2-full/drivers/scsi/tmscsim.h	2003-09-02 20:31:21.000000000 +0200
@@ -11,7 +11,7 @@
 #include <linux/types.h>
 #include <linux/config.h>
 
-#define IRQ_NONE 255
+#define SCSI_IRQ_NONE 255
 
 #define MAX_ADAPTER_NUM 	4
 #define MAX_SG_LIST_BUF 	16	/* Not used */

Re: [2.4 patch] Fix IRQ_NONE clash in SCSI drivers

[Arnaldo Carvalho de Melo]

Em Tue, Sep 02, 2003 at 08:44:36PM +0200, Adrian Bunk escreveu:
> On Wed, Aug 27, 2003 at 02:52:45PM -0300, Marcelo Tosatti wrote:
> >...
> > Summary of changes from v2.4.22 to v2.4.23-pre1
> > ============================================
> >...
> > Arnaldo Carvalho de Melo:
> >   o irqreturn_t compatibility with 2.6
> >...
> 
> This change added an (empty) IRQ_NONE #define to interrupt.h.
> 
> Several scsi drivers are already using an IRQ_NONE.  Rename that to
> SCSI_IRQ_NONE (a similar change was done in 2.5 by Andrew Morton several
> months ago).
> 
> I've tested the compilation with 2.4.23-pre2.

Thanks a lot!

- Arnaldo

Re: [2.4 patch] Fix IRQ_NONE clash in SCSI drivers

[Marc-Christian Petersen]

[-- Attachment #1: Type: text/plain, Size: 400 bytes --]

On Tuesday 02 September 2003 20:44, Adrian Bunk wrote:

Hi Adrian,

> This change added an (empty) IRQ_NONE #define to interrupt.h.
> Several scsi drivers are already using an IRQ_NONE.  Rename that to
> SCSI_IRQ_NONE (a similar change was done in 2.5 by Andrew Morton several
> months ago).

right, but you forgot one header :-) ... Attached is a patch.

Marcelo, please apply this too.

ciao, Marc

[-- Attachment #2: irqreturnt-compatibility-2.6-missings.patch --]
[-- Type: text/x-diff, Size: 296 bytes --]

--- a/drivers/scsi/NCR5380.h	2001-12-21 18:41:55.000000000 +0100
+++ b/drivers/scsi/NCR5380.h	2003-09-03 10:45:57.000000000 +0200
@@ -233,7 +233,7 @@
  * Scsi_Host structure
  */
 
-#define IRQ_NONE	255
+#define SCSI_IRQ_NONE	255
 #define DMA_NONE	255
 #define IRQ_AUTO	254
 #define DMA_AUTO	254

Re: [2.4 patch] Fix IRQ_NONE clash in SCSI drivers

[Adrian Bunk]

On Wed, Sep 03, 2003 at 10:48:53AM +0200, Marc-Christian Petersen wrote:
> On Tuesday 02 September 2003 20:44, Adrian Bunk wrote:
> 
> Hi Adrian,
> 
> > This change added an (empty) IRQ_NONE #define to interrupt.h.
> > Several scsi drivers are already using an IRQ_NONE.  Rename that to
> > SCSI_IRQ_NONE (a similar change was done in 2.5 by Andrew Morton several
> > months ago).
> 
> right, but you forgot one header :-) ... Attached is a patch.

Uups, I fixed it but forgot to include it in the patch.

Thanks for noticing it!

> Marcelo, please apply this too.
> 
> ciao, Marc

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed