Skip to contentSkip to conversation

2.6.10-kj

announce New Domen Puncer 7 participants 11 messages Dec 25, 2004 – Jan 19, 2005

Subsystems (2)
Kernel janitors ATM
lore
Share & tools
lore

Domen Puncer

author
#0 lore raw

Hi.

Santa brought another present :-)

I'll start mailing new patches these days, and after external trees get merged, I'll be bugging you with the old ones.

Patchset is at http://coderock.org/kj/2.6.10-kj/

new in this release:
--------------------
comment-drivers_block_floppy.c.patch
From: [email protected]
Subject: [KJ] [PATCH][RESEND] floppy: relocate devfs comment

module_param-drivers_net_3c59x.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch] net/3c59x: module_param conversions

schedule_cleanup-drivers_usb_class_usblp.c.patch
From: Nishanth Aravamudan <[email protected]>
Subject: [KJ] [PATCH] class/usblp: cleanup usblp_write()

typo-sound_isa_es18xx.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch 8/8] ifdef typos: sound_isa_es18xx.c

typos-arch_ppc_platforms_prep_setup.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] Re: ifdef typos: arch_ppc_platforms_prep_setup.c -another one

typo-arch_ppc_syslib_ppc4xx_dma.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch 2/8] ifdef typos: arch_ppc_syslib_ppc4xx_dma.c

typo-arch_sh_boards_renesas_hs7751rvoip_io.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch 3/8] ifdef typos: arch_sh_boards_renesas_hs7751rvoip_io.c

typo-drivers_char_ipmi_ipmi_si_intf.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch 4/8] ifdef typos: drivers_char_ipmi_ipmi_si_intf.c

typo-drivers_net_wireless_wavelan_cs.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch 5/8] ifdef typos: drivers_net_wireless_wavelan_cs.c

typo-drivers_usb_net_usbnet.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch 6/8] ifdef typos: drivers_usb_net_usbnet.c

typo-sound_isa_cs423x_cs4231_lib.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch 7/8] ifdef typos: sound_isa_cs423x_cs4231_lib.c

typo_au1000-arch_mips.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch] mips: AU1[0X]00_USB_DEVICE typo

typo-arch_ppc64_kernel_rtasd.c.patch
From: Domen Puncer <[email protected]>
Subject: [KJ] [patch] ppc64: semicolon in rtasd.c

remove_file-arch_alpha_lib_dbg_stackcheck.S.patch
delete unused file

remove_file-arch_alpha_lib_dbg_stackkill.S.patch
delete unused file

remove_file-arch_alpha_lib_stacktrace.c.patch
delete unused file

remove_file-arch_arm26_boot_compressed_hw_bse.c.patch
delete unused file

remove_file-arch_m68k_apollo_dn_debug.c.patch
delete unused file

remove_file-arch_m68k_sun3x_sun3x_ksyms.c.patch
delete unused file

remove_file-arch_m68knommu_platform_68EZ328_ucsimm_crt0_himem.S.patch
delete unused file

remove_file-arch_m68knommu_platform_68VZ328_ucdimm_crt0_himem.S.patch
delete unused file

remove_file-arch_mips_arc_salone.c.patch
delete unused file

remove_file-arch_mips_pmc_sierra_yosemite_ht_irq.c.patch
delete unused file

remove_file-arch_ppc64_boot_no_initrd.c.patch
delete unused file

remove_file-arch_ppc_kernel_find_name.c.patch
delete unused file

remove_file-arch_ppc_syslib_ppc4xx_pm.c.patch
delete unused file

remove_file-arch_ppc_syslib_ppc4xx_serial.c.patch
delete unused file

remove_file-arch_sh64_lib_old_checksum.c.patch
delete unused file

remove_file-arch_um_include_umn.h.patch
delete unused file

remove_file-arch_x86_64_lib_old_checksum.c.patch
delete unused file

remove_file-drivers_char_hp600_keyb.c.patch
delete unused file

remove_file-drivers_char_rio_cdproto.h.patch
delete unused file

remove_file-drivers_char_rsf16fmi.h.patch
delete unused file

remove_file-drivers_net_gt64240eth.h.patch
delete unused file

remove_file-drivers_parport_parport_arc.c.patch
delete unused file

remove_file-drivers_pcmcia_au1000_generic.c.patch
delete unused file

remove_file-drivers_pcmcia_au1000_pb1x00.c.patch
delete unused file

remove_file-drivers_scsi_dpt_dpt_osdutil.h.patch
delete unused file

remove_file-drivers_video_aty_xlinit.c.patch
delete unused file

remove_file-drivers_video_riva_nv4ref.h.patch
delete unused file

remove_file-fs_jffs2_histo.h.patch
delete unused file

remove_file-fs_jfs_jfs_defragfs.h.patch
delete unused file

remove_file-include_asm_alpha_numnodes.h.patch
delete unused file

remove_file-include_asm_arm26_ian_char.h.patch
delete unused file

remove_file-include_asm_arm_arch_epxa10db_mode_ctrl00.h.patch
delete unused file

remove_file-include_asm_arm_arch_epxa10db_pld_conf00.h.patch
delete unused file

remove_file-include_asm_arm_hardware_linkup_l1110.h.patch
delete unused file

remove_file-include_asm_m68k_atari_SCCserial.h.patch
delete unused file

remove_file-include_asm_m68knommu_io_hw_swap.h.patch
delete unused file

remove_file-include_asm_m68knommu_semp3.h.patch
delete unused file

remove_file-include_asm_mips_gfx.h.patch
delete unused file

remove_file-include_asm_mips_it8172_it8172_lpc.h.patch
delete unused file

remove_file-include_asm_mips_mach_au1x00_au1100_mmc.h.patch
delete unused file

remove_file-include_asm_mips_mipsprom.h.patch
delete unused file

remove_file-include_asm_mips_ng1.h.patch
delete unused file

remove_file-include_asm_mips_ng1hw.h.patch
delete unused file

remove_file-include_asm_mips_riscos_syscall.h.patch
delete unused file

remove_file-include_asm_parisc_bootdata.h.patch
delete unused file

remove_file-include_asm_ppc64_iSeries_iSeries_fixup.h.patch
delete unused file

remove_file-include_asm_um_arch_signal_i386.h.patch
delete unused file

remove_file-include_asm_um_archparam_i386.h.patch
delete unused file

remove_file-include_asm_um_archparam_ppc.h.patch
delete unused file

remove_file-include_asm_um_module_generic.h.patch
delete unused file

remove_file-include_asm_um_module_i386.h.patch
delete unused file

remove_file-include_asm_um_processor_i386.h.patch
delete unused file

remove_file-include_asm_um_processor_ppc.h.patch
delete unused file

remove_file-include_asm_um_ptrace_i386.h.patch
delete unused file

remove_file-include_asm_um_sigcontext_i386.h.patch
delete unused file

remove_file-include_asm_um_sigcontext_ppc.h.patch
delete unused file

remove_file-include_asm_um_system_i386.h.patch
delete unused file

remove_file-include_asm_um_system_ppc.h.patch
delete unused file

remove_file-include_linux_byteorder_pdp_endian.h.patch
delete unused file

remove_file-include_linux_netfilter_ipv4_ip_logging.h.patch
delete unused file

remove_file-include_linux_netfilter_ipv6_ip6_logging.h.patch
delete unused file

remove_file-include_linux_umsdos_fs_i.h.patch
delete unused file

remove_file-include_sound_soundmem.h.patch
delete unused file

remove_file-net_sunrpc_auth_gss_gss_pseudoflavors.c.patch
delete unused file

remove_file-net_sunrpc_auth_gss_sunrpcgss_syms.c.patch
delete unused file

remove_file-sound_core_seq_oss_seq_oss_misc.c.patch
delete unused file

remove_file-sound_oss_maestro_tables.h.patch
delete unused file

remove_file-sound_pci_cs46xx_imgs_cwcemb80.h.patch
delete unused file

cleanup-drivers_media_radio_radio-sf16fmi.c.patch
From: [email protected] (Marcel Sebek)
Subject: [KJ] [PATCH] radio-sf16fmi cleanup

cleanup-drivers_media_radio_miropcm20-radio.c.patch
From: [email protected] (Marcel Sebek)
Subject: [KJ] [PATCH] miropcm20-radio cleanup

cleanup-drivers_atm_he.c.patch
From: [email protected] (Marcel Sebek)
Subject: [KJ] [PATCH] drivers/atm/he.c remove dead code

array_size-fs_proc_base.c.bak.patch
From: walter harms <[email protected]>
Subject: [KJ] inux-2.6.9/fs/proc/base.c: array size

remove_sprintf-fs_proc_proc_tty.c.bak.patch
From: walter harms <[email protected]>
Subject: [KJ] linux-2.6.9/fs/proc/proc_tty.c: avoid array

kj_tag.patch
-kj

merged:
-------
min-max-arch_mips_au1000_common_usbdev
msleep-drivers_net_irda_pcmcia_yenta_socket
check-register_netdevice_notifier-net_8021q_vlan
check-register_netdevice_notifier-net_atm_mpc

changed:
--------
min-max-arch_sh_boards_bigsur_io.patch
changed to min_t
ssleep-drivers_scsi_qla2xxx_qla_os
changed to msleep
msleep_interruptible-drivers_macintosh_mediabay.patch
fixed typo

dropped:
--------
remove-pci-find-device-arch_mips_pci_pci-hplj
file removed
fix-test-kernel_sched
nacked

all patches:
------------
remove-old-ifdefs-aic7xxx_osm_pci.patch min-max-ide_ide-timing.h.patch list-for-each-entry-drivers_chan_kern.patch list-for-each-entry-drivers_macintosh_via-pmu.patch list-for-each-entry-drivers_net_ppp_generic.patch list-for-each-entry-fs_jffs_intrep.patch list-for-each-entry-fs_namespace.patch list-for-each-entry-safe-arch_i386_mm_pageattr.patch list-for-each-entry-safe-fs_coda_psdev.patch list-for-each-entry-safe-fs_dquot.patch list-for-each-fs_dcache.patch msleep-drivers_ide_ide-tape.patch msleep-drivers_ieee1394_sbp2.patch pr_debug-drivers_block_umem.patch list-for-each-drivers_net_ipv6_ip6_fib.patch list-for-each-drivers_net_tulip_de4x5.patch min-max-arch_sh_boards_bigsur_io.patch min-max-arch_sh_cchips_hd6446x_hd64465_io.patch msleep-drivers_atm_ambassador.patch msleep-drivers_block_xd.patch msleep-drivers_ide_ide-cd.patch msleep-drivers_ide_ide-cs.patch msleep-drivers_mtd_chips_amd_flash.patch msleep-drivers_mtd_chips_cfi_cmdset_0001.patch msleep-drivers_mtd_chips_cfi_cmdset_0002.patch msleep-drivers_mtd_chips_cfi_cmdset_0020.patch msleep_interruptible-drivers_base_dmapool.patch msleep_interruptible-drivers_block_cciss.patch msleep_interruptible-drivers_block_pcd.patch msleep_interruptible-drivers_block_pf.patch msleep_interruptible-drivers_block_pg.patch msleep_interruptible-drivers_block_pt.patch msleep_interruptible-drivers_cdrom_sonycd535.patch msleep_interruptible-drivers_macintosh_mediabay.patch set_current_state-drivers_block_swim3.patch set_current_state-drivers_block_swim_iop.patch add_module_version-drivers_net_8139cp.patch for-each-pci-dev-arch_i386_pci_acpi.patch for-each-pci-dev-arch_i386_pci_i386.patch for-each-pci-dev-drivers_char_agp_generic.patch for-each-pci-dev-drivers_char_agp_isoch.patch function-string-arch-mips.patch msleep-drivers_media_radio_radio-zoltrix.patch msleep-drivers_net_3c505.patch msleep-drivers_net_e1000_e1000_osdep.patch msleep-drivers_net_irda_act200l-sir.patch msleep-drivers_net_irda_irtty-sir.patch msleep-drivers_net_irda_ma600-sir.patch msleep-drivers_net_irda_pcmcia_xirc2ps_cs.patch msleep-drivers_net_irda_sir_dev.patch msleep-drivers_net_ixgb_ixgb_osdep.patch msleep-drivers_net_ni65.patch msleep-drivers_net_ns83820.patch msleep-drivers_net_s2io.patch msleep-drivers_net_wan_cosa.patch msleep-drivers_net_wireless_prism54_islpci_dev.patch msleep-drivers_scsi_ibmvscsi.patch msleep-drivers_scsi_ide-scsi.patch msleep-drivers_scsi_imm.patch msleep-drivers_scsi_osst.patch msleep-drivers_scsi_ppa.patch msleep-drivers_scsi_qla1280.patch msleep_interruptible-drivers_net_cs89x0.patch msleep_interruptible-drivers_net_e1000_e1000_ethtool.patch msleep_interruptible-drivers_net_gt96100eth.patch msleep_interruptible-drivers_net_irda_tekram-sir.patch msleep_interruptible-drivers_net_ixgb_ixgb_ethtool.patch msleep_interruptible-drivers_net_pcnet32.patch msleep_interruptible-drivers_net_sb1000.patch msleep_interruptible-drivers_net_slip.patch msleep_interruptible-drivers_net_wan_cycx_drv.patch msleep_interruptible-drivers_parport_ieee1284_ops.patch msleep_interruptible-drivers_parport_ieee1284.patch msleep_interruptible-drivers_parport_parport_pc.patch msleep_interruptible-drivers_s390_net_ctctty.patch msleep_interruptible-drivers_sbus_char_aurora.patch msleep_interruptible-drivers_sbus_char_envctrl.patch msleep_interruptible-drivers_scsi_dpt_i2o.patch msleep_interruptible-drivers_scsi_st.patch msleep_interruptible-drivers_tc_zs.patch msleep+msleep_interruptible-drivers_net_e100.patch msleep+msleep_interruptible-drivers_net_tokenring_ibmtr.patch msleep+msleep_interruptible-drivers_net_wireless_airo.patch msleep+ssleep-drivers_net_appletalk_ltpc.patch pci_dev_present-arch_ia64_hp_common_sba_iommu.patch pci_dev_present-arch_ia64_pci_pci.patch pci_dev_present-drivers_ide_pci_alim15x3.patch pci_get_device-drivers_char_agp_amd64-agp.patch pci_get_device-drivers_char_agp_intel-agp.patch pci_get_device-drivers_char_agp_intel-mch-agp.patch remove-pci-find-device-arch_sparc64_kernel_ebus.patch remove-pci-find-device-drivers_net_e1000_e1000_main.patch remove-pci-find-device-drivers_net_gt96100eth.patch remove-pci-find-device-drivers_net_ixgb_ixgb_main.patch remove-pci-find-device-drivers_net_tg3.patch set_current_state-drivers_net_irda_stir4200.patch set_current_state-drivers_net_tokenring_tms380tr.patch set_current_state-drivers_net_wan_farsync.patch ssleep-drivers_net_wireless_orinoco_plx.patch ssleep-drivers_net_wireless_orinoco_tmd.patch ssleep+msleep_interruptible-drivers_net_tokenring_lanstreamer.patch fix-comment-fs_jbd_journal.patch lib-parser-fs_devpts_inode.patch msleep_interruptible-drivers_cdrom_sonycd535_2.patch msleep_interruptible-drivers_message_fusion_mptbase.patch msleep_interruptible-drivers_net_ewrk3.patch reorder-state-drivers_char_snsc.patch reorder-state-drivers_video_pxafb.patch reorder-state-drivers_video_sa1100fb.patch set_current_state-drivers_input_joystick_iforce_iforce-packets.patch ssleep-drivers_scsi_qla2xxx_qla_os.patch docs-fs_super.patch
docs-kernel_sysctl.patch
kconfig-arch_sh_drivers_pci.patch kconfig-arch_sparc64.patch
module_parm-net_ne2k-pci.patch module_parm-net_wireless_atmel_cs.patch module_parm-net_wireless_orinoco.patch module_parm-net_wireless_wavelan.p.h.patch module_parm-net_wireless_wl3501_cs.patch msleep-sound_sparc_cs4231.patch
myri_code_cleanup.patch
printk-drivers-scsi-zalon.patch strlcpy-net_wireless_wavelan.patch comment-drivers_block_floppy.c.patch module_param-drivers_net_3c59x.c.patch schedule_cleanup-drivers_usb_class_usblp.c.patch typo-sound_isa_es18xx.c.patch typos-arch_ppc_platforms_prep_setup.c.patch typo-arch_ppc_syslib_ppc4xx_dma.c.patch typo-arch_sh_boards_renesas_hs7751rvoip_io.c.patch typo-drivers_char_ipmi_ipmi_si_intf.c.patch typo-drivers_net_wireless_wavelan_cs.c.patch typo-drivers_usb_net_usbnet.c.patch typo-sound_isa_cs423x_cs4231_lib.c.patch typo_au1000-arch_mips.patch typo-arch_ppc64_kernel_rtasd.c.patch remove_file-arch_alpha_lib_dbg_stackcheck.S.patch remove_file-arch_alpha_lib_dbg_stackkill.S.patch remove_file-arch_alpha_lib_stacktrace.c.patch remove_file-arch_arm26_boot_compressed_hw_bse.c.patch remove_file-arch_m68k_apollo_dn_debug.c.patch remove_file-arch_m68k_sun3x_sun3x_ksyms.c.patch remove_file-arch_m68knommu_platform_68EZ328_ucsimm_crt0_himem.S.patch remove_file-arch_m68knommu_platform_68VZ328_ucdimm_crt0_himem.S.patch remove_file-arch_mips_arc_salone.c.patch remove_file-arch_mips_pmc_sierra_yosemite_ht_irq.c.patch remove_file-arch_ppc64_boot_no_initrd.c.patch remove_file-arch_ppc_kernel_find_name.c.patch remove_file-arch_ppc_syslib_ppc4xx_pm.c.patch remove_file-arch_ppc_syslib_ppc4xx_serial.c.patch remove_file-arch_sh64_lib_old_checksum.c.patch remove_file-arch_um_include_umn.h.patch remove_file-arch_x86_64_lib_old_checksum.c.patch remove_file-drivers_char_hp600_keyb.c.patch remove_file-drivers_char_rio_cdproto.h.patch remove_file-drivers_char_rsf16fmi.h.patch remove_file-drivers_net_gt64240eth.h.patch remove_file-drivers_parport_parport_arc.c.patch remove_file-drivers_pcmcia_au1000_generic.c.patch remove_file-drivers_pcmcia_au1000_pb1x00.c.patch remove_file-drivers_scsi_dpt_dpt_osdutil.h.patch remove_file-drivers_video_aty_xlinit.c.patch remove_file-drivers_video_riva_nv4ref.h.patch remove_file-fs_jffs2_histo.h.patch remove_file-fs_jfs_jfs_defragfs.h.patch remove_file-include_asm_alpha_numnodes.h.patch remove_file-include_asm_arm26_ian_char.h.patch remove_file-include_asm_arm_arch_epxa10db_mode_ctrl00.h.patch remove_file-include_asm_arm_arch_epxa10db_pld_conf00.h.patch remove_file-include_asm_arm_hardware_linkup_l1110.h.patch remove_file-include_asm_m68k_atari_SCCserial.h.patch remove_file-include_asm_m68knommu_io_hw_swap.h.patch remove_file-include_asm_m68knommu_semp3.h.patch remove_file-include_asm_mips_gfx.h.patch remove_file-include_asm_mips_it8172_it8172_lpc.h.patch remove_file-include_asm_mips_mach_au1x00_au1100_mmc.h.patch remove_file-include_asm_mips_mipsprom.h.patch remove_file-include_asm_mips_ng1.h.patch remove_file-include_asm_mips_ng1hw.h.patch remove_file-include_asm_mips_riscos_syscall.h.patch remove_file-include_asm_parisc_bootdata.h.patch remove_file-include_asm_ppc64_iSeries_iSeries_fixup.h.patch remove_file-include_asm_um_arch_signal_i386.h.patch remove_file-include_asm_um_archparam_i386.h.patch remove_file-include_asm_um_archparam_ppc.h.patch remove_file-include_asm_um_module_generic.h.patch remove_file-include_asm_um_module_i386.h.patch remove_file-include_asm_um_processor_i386.h.patch remove_file-include_asm_um_processor_ppc.h.patch remove_file-include_asm_um_ptrace_i386.h.patch remove_file-include_asm_um_sigcontext_i386.h.patch remove_file-include_asm_um_sigcontext_ppc.h.patch remove_file-include_asm_um_system_i386.h.patch remove_file-include_asm_um_system_ppc.h.patch remove_file-include_linux_byteorder_pdp_endian.h.patch remove_file-include_linux_netfilter_ipv4_ip_logging.h.patch remove_file-include_linux_netfilter_ipv6_ip6_logging.h.patch remove_file-include_linux_umsdos_fs_i.h.patch remove_file-include_sound_soundmem.h.patch remove_file-net_sunrpc_auth_gss_gss_pseudoflavors.c.patch remove_file-net_sunrpc_auth_gss_sunrpcgss_syms.c.patch remove_file-sound_core_seq_oss_seq_oss_misc.c.patch remove_file-sound_oss_maestro_tables.h.patch remove_file-sound_pci_cs46xx_imgs_cwcemb80.h.patch cleanup-drivers_media_radio_radio-sf16fmi.c.patch cleanup-drivers_media_radio_miropcm20-radio.c.patch cleanup-drivers_atm_he.c.patch array_size-fs_proc_base.c.bak.patch remove_sprintf-fs_proc_proc_tty.c.bak.patch
kj_tag.patch

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

#1 lore raw

PATCH 1/1 atm/ambassador: use msleep() instead of schedule_timeout()

Quoted 6 lines · Domen Puncer

Hi.

Santa brought another present :-)

I'll start mailing new patches these days, and after external trees get merged, I'll be bugging you with the old ones.

<snip>

Quoted 2 lines

all patches:
------------

<snip>

Quoted 1 line

msleep-drivers_atm_ambassador.patch

Please consider updating to the following patch.

Description: Multiple schedule_timeout() related fixes. Generally uses msleep() to guarantee the task delays as expected. In one place, this allowed for the deletion of a variable. In a few places, I reverted back to using TASK_INTERRUPTIBLE, because the code makes little sense otherwise. In these places, the units of the timeout variable has been changed from jiffies to msecs. As far as I can see, this driver is still not fixed, as there is no response to the signal which may have interrupted the sleep. But this patch at least brings it closer.

Signed-off-by Nishanth Aravamudan <[email protected]>
drivers/atm/ambassador.c +14 -29
@@ -574,7 +574,6 @@ static int command_do (amb_dev * dev, co
   amb_cq * cq = &dev->cq;
   volatile amb_cq_ptrs * ptrs = &cq->ptrs;
   command * my_slot;
-  unsigned long timeout;
   
   PRINTD (DBG_FLOW|DBG_CMD, "command_do %p", dev);
   

@@ -599,20 +598,14 @@ static int command_do (amb_dev * dev, co
     // mail the command
     wr_mem (dev, offsetof(amb_mem, mb.adapter.cmd_address), virt_to_bus (ptrs->in));
     
-    // prepare to wait for cq->pending milliseconds
-    // effectively one centisecond on i386
-    timeout = (cq->pending*HZ+999)/1000;
-    
     if (cq->pending > cq->high)
       cq->high = cq->pending;
     spin_unlock (&cq->lock);
     
-    while (timeout) {
-      // go to sleep
-      // PRINTD (DBG_CMD, "wait: sleeping %lu for command", timeout);
-      set_current_state(TASK_UNINTERRUPTIBLE);
-      timeout = schedule_timeout (timeout);
-    }
+    // these comments were in a while-loop before, msleep removes the loop
+    // go to sleep
+    // PRINTD (DBG_CMD, "wait: sleeping %lu for command", timeout);
+    msleep(cq->pending);
     
     // wait for my slot to be reached (all waiters are here or above, until...)
     while (ptrs->out != my_slot) {

@@ -1799,12 +1792,11 @@ static int __init do_loader_command (vol
   // dump_loader_block (lb);
   wr_mem (dev, offsetof(amb_mem, doorbell), virt_to_bus (lb) & ~onegigmask);
   
-  timeout = command_timeouts[cmd] * HZ/100;
+  timeout = command_timeouts[cmd] * 10;
   
   while (!lb->result || lb->result == cpu_to_be32 (COMMAND_IN_PROGRESS))
     if (timeout) {
-      set_current_state(TASK_UNINTERRUPTIBLE);
-      timeout = schedule_timeout (timeout);
+      timeout = msleep_interruptible(timeout);
     } else {
       PRINTD (DBG_LOAD|DBG_ERR, "command %d timed out", cmd);
       dump_registers (dev);

@@ -1814,10 +1806,10 @@ static int __init do_loader_command (vol
   
   if (cmd == adapter_start) {
     // wait for start command to acknowledge...
-    timeout = HZ/10;
+    timeout = 100;
     while (rd_plain (dev, offsetof(amb_mem, doorbell)))
       if (timeout) {
-	timeout = schedule_timeout (timeout);
+	timeout = msleep_interruptible(timeout);
       } else {
 	PRINTD (DBG_LOAD|DBG_ERR, "start command did not clear doorbell, res=%08x",
 		be32_to_cpu (lb->result));

@@ -1932,17 +1924,12 @@ static int amb_reset (amb_dev * dev, int
   if (diags) { 
     unsigned long timeout;
     // 4.2 second wait
-    timeout = HZ*42/10;
-    while (timeout) {
-      set_current_state(TASK_UNINTERRUPTIBLE);
-      timeout = schedule_timeout (timeout);
-    }
+    msleep(4200);
     // half second time-out
-    timeout = HZ/2;
+    timeout = 500;
     while (!rd_plain (dev, offsetof(amb_mem, mb.loader.ready)))
       if (timeout) {
-        set_current_state(TASK_UNINTERRUPTIBLE);
-	timeout = schedule_timeout (timeout);
+        timeout = msleep_interruptible(timeout);
       } else {
 	PRINTD (DBG_LOAD|DBG_ERR, "reset timed out");
 	return -ETIMEDOUT;

@@ -2056,14 +2043,12 @@ static int __init amb_talk (amb_dev * de
   wr_mem (dev, offsetof(amb_mem, doorbell), virt_to_bus (&a));
   
   // 2.2 second wait (must not touch doorbell during 2 second DMA test)
-  timeout = HZ*22/10;
-  while (timeout)
-    timeout = schedule_timeout (timeout);
+  msleep(2200);
   // give the adapter another half second?
-  timeout = HZ/2;
+  timeout = 500;
   while (rd_plain (dev, offsetof(amb_mem, doorbell)))
     if (timeout) {
-      timeout = schedule_timeout (timeout);
+      timeout = msleep_interruptible(timeout);
     } else {
       PRINTD (DBG_INIT|DBG_ERR, "adapter init timed out");
       return -ETIMEDOUT;

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

#2 lore raw

PATCH 1/1 ide/ide-cd: use ssleep() instead of schedule_timeout()

In reply to Domen Puncer
Quoted 9 lines · Domen Puncer

Hi.

Santa brought another present :-)

I'll start mailing new patches these days, and after external trees get merged, I'll be bugging you with the old ones.

Patchset is at http://coderock.org/kj/2.6.10-kj/

<snip>

Quoted 1 line

msleep-drivers_ide_ide-cd.patch

Please consider updating to the following patch:

Description: Uses ssleep() in place of cdrom_sleep() to guarantee the task delays as expected. Remove cdrom_sleep() definition, as this is the only place where it is called.

Signed-off-by Nishanth Aravamudan
drivers/ide/ide-cd.c +1 -14
@@ -1520,19 +1520,6 @@ static ide_startstop_t cdrom_do_packet_c
 }
 
 
-/* Sleep for TIME jiffies.
-   Not to be called from an interrupt handler. */
-static
-void cdrom_sleep (int time)
-{
-	int sleep = time;
-
-	do {
-		set_current_state(TASK_INTERRUPTIBLE);
-		sleep = schedule_timeout(sleep);
-	} while (sleep);
-}
-
 static
 int cdrom_queue_packet_command(ide_drive_t *drive, struct request *rq)
 {

@@ -1567,7 +1554,7 @@ int cdrom_queue_packet_command(ide_drive
 				/* The drive is in the process of loading
 				   a disk.  Retry, but wait a little to give
 				   the drive time to complete the load. */
-				cdrom_sleep(2 * HZ);
+				ssleep(2);
 			} else {
 				/* Otherwise, don't retry. */
 				retries = 0;

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

Acked-by on v1 1/1
#3 lore raw

Re: [UPDATE PATCH] ide/ide-cd: use ssleep() instead of schedule_timeout()

Quoted 15 lines · Nishanth Aravamudan
Quoted 9 lines · Domen Puncer

Hi.

Santa brought another present :-)

I'll start mailing new patches these days, and after external trees get merged, I'll be bugging you with the old ones.

Patchset is at http://coderock.org/kj/2.6.10-kj/

<snip>

Quoted 1 line

msleep-drivers_ide_ide-cd.patch

Quoted 7 lines

Please consider updating to the following patch:

Description: Uses ssleep() in place of cdrom_sleep() to guarantee the task delays as expected. Remove cdrom_sleep() definition, as this is the only place where it is called.

Signed-off-by Nishanth Aravamudan

Looks fine.

Acked-by Jens Axboe <[email protected]>
Quoted 32 lines
drivers/ide/ide-cd.c +1 -14
@@ -1520,19 +1520,6 @@ static ide_startstop_t cdrom_do_packet_c
 }
 
 
-/* Sleep for TIME jiffies.
-   Not to be called from an interrupt handler. */
-static
-void cdrom_sleep (int time)
-{
-	int sleep = time;
-
-	do {
-		set_current_state(TASK_INTERRUPTIBLE);
-		sleep = schedule_timeout(sleep);
-	} while (sleep);
-}
-
 static
 int cdrom_queue_packet_command(ide_drive_t *drive, struct request *rq)
 {

@@ -1567,7 +1554,7 @@ int cdrom_queue_packet_command(ide_drive
 				/* The drive is in the process of loading
 				   a disk.  Retry, but wait a little to give
 				   the drive time to complete the load. */
-				cdrom_sleep(2 * HZ);
+				ssleep(2);
 			} else {
 				/* Otherwise, don't retry. */
 				retries = 0;
Signature · 1 line
Jens Axboe

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

#5 lore raw

PATCH 1/1 ieee1394/sbp2: use ssleep() instead of schedule_timeout()

In reply to Domen Puncer
Quoted 9 lines · Domen Puncer

Hi.

Santa brought another present :-)

I'll start mailing new patches these days, and after external trees get merged, I'll be bugging you with the old ones.

Patchset is at http://coderock.org/kj/2.6.10-kj/

<snip>

Quoted 2 lines

all patches:
------------

<snip>

Quoted 1 line

msleep-drivers_ieee1394_sbp2.patch

Please consider updating to the following patch:

Description: Use ssleep() instead of schedule_timeout() to guarantee the task delays as expected. The existing code should not really need to run in TASK_INTERRUPTIBLE, as there is no check for signals (or even an early return value whatsoever). ssleep() takes care of these issues.

Signed-off-by Nishanth Aravamudan <[email protected]>
drivers/ieee1394/sbp2.c +1 -2
@@ -902,8 +902,7 @@ alloc_fail:
 	 * connected to the sbp2 device being removed. That host would
 	 * have a certain amount of time to relogin before the sbp2 device
 	 * allows someone else to login instead. One second makes sense. */
-	set_current_state(TASK_INTERRUPTIBLE);
-	schedule_timeout(HZ);
+	ssleep(1);
 
 	/*
 	 * Login to the sbp-2 device

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

#6 lore raw

Re: [UPDATE PATCH] ieee1394/sbp2: use ssleep() instead of schedule_timeout()

Quoted 9 lines · Nishanth Aravamudan

Description: Use ssleep() instead of schedule_timeout() to guarantee the task delays as expected. The existing code should not really need to run in TASK_INTERRUPTIBLE, as there is no check for signals (or even an early return value whatsoever). ssleep() takes care of these issues.

--- 2.6.10-v/drivers/ieee1394/sbp2.c 2004-12-24 13:34:00.000000000
-0800
+++ 2.6.10/drivers/ieee1394/sbp2.c 2005-01-05 14:23:05.000000000 -0800

drivers/ieee1394/sbp2.c alloc_fail: · line 905

Quoted 6 lines
drivers/ieee1394/sbp2.c +1 -2
@@ -902,8 +902,7 @@ alloc_fail:
 	 * connected to the sbp2 device being removed. That host would
 	 * have a certain amount of time to relogin before the sbp2 device
 	 * allows someone else to login instead. One second makes sense. */
-	set_current_state(TASK_INTERRUPTIBLE);
-	schedule_timeout(HZ);
+	ssleep(1);

Maybe the current code is _deliberately_ accepting interruption by signals but trying to complete sbp2_probe() anyway. However it seems more plausible to me to abort the device probe, for example like this:

if (msleep_interruptible(1000)) {
	sbp2_remove_device(scsi_id);
	return -EINTR;
}

Anyway, signal handling does not appear to be critical there.

Signature · 3 lines
Stefan Richter
-=====-=-=-= ---= -=--=
http://arcgraph.de/sr/

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

#7 lore raw

Re: [UPDATE PATCH] ieee1394/sbp2: use ssleep() instead of schedule_timeout()

Quoted 22 lines · Stefan Richter
Quoted 11 lines · Nishanth Aravamudan

Description: Use ssleep() instead of schedule_timeout() to guarantee the task
delays as expected. The existing code should not really need to run in TASK_INTERRUPTIBLE, as there is no check for signals (or even an early return
value whatsoever). ssleep() takes care of these issues.

--- 2.6.10-v/drivers/ieee1394/sbp2.c 2004-12-24 13:34:00.000000000
-0800
+++ 2.6.10/drivers/ieee1394/sbp2.c 2005-01-05 14:23:05.000000000 -0800

drivers/ieee1394/sbp2.c alloc_fail: · line 905

Quoted 6 lines
drivers/ieee1394/sbp2.c +1 -2
@@ -902,8 +902,7 @@ alloc_fail:
 	 * connected to the sbp2 device being removed. That host would
 	 * have a certain amount of time to relogin before the sbp2 device
 	 * allows someone else to login instead. One second makes sense. */
-	set_current_state(TASK_INTERRUPTIBLE);
-	schedule_timeout(HZ);
+	ssleep(1);

Maybe the current code is _deliberately_ accepting interruption by signals but trying to complete sbp2_probe() anyway. However it seems more plausible to me to abort the device probe, for example like this:

Quoted 4 lines
if (msleep_interruptible(1000)) {
	sbp2_remove_device(scsi_id);
	return -EINTR;
}

You might be right, but I'd like to get Ben's input on this, as I honeslty am unsure. To be fair, I am trying to audit all usage of schedule_timeout() and the semantic interpretation (to me) of using TASK_INTERRUPTIBLE is that you wish to sleep a certain amount of time, but also are prepared for an early return on either signals or wait-queue events. msleep_interruptible() cleanly removes this second issue, but still requires the caller to respond appropriately if there is a return value. Hence, I like your change. I think it makes the most sense. Since I didn't/don't know how the device works, I was not able to make the change myself. Thanks for your input!

Quoted 1 line

Anyway, signal handling does not appear to be critical there.

Just out of curiousity, doesn't that run the risk, though, of signal_pending(current) being true for quite a bit of time following the timeout?

Thanks,
Nish

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

#8 lore raw

Re: [UPDATE PATCH] ieee1394/sbp2: use ssleep() instead of schedule_timeout()

Quoted 28 lines · Nishanth Aravamudan
Quoted 22 lines · Stefan Richter
Quoted 11 lines · Nishanth Aravamudan

Description: Use ssleep() instead of schedule_timeout() to guarantee the task
delays as expected. The existing code should not really need to run in TASK_INTERRUPTIBLE, as there is no check for signals (or even an early return
value whatsoever). ssleep() takes care of these issues.

--- 2.6.10-v/drivers/ieee1394/sbp2.c 2004-12-24 13:34:00.000000000
-0800
+++ 2.6.10/drivers/ieee1394/sbp2.c 2005-01-05 14:23:05.000000000 -0800

drivers/ieee1394/sbp2.c alloc_fail: · line 905

Quoted 6 lines
drivers/ieee1394/sbp2.c +1 -2
@@ -902,8 +902,7 @@ alloc_fail:
 	 * connected to the sbp2 device being removed. That host would
 	 * have a certain amount of time to relogin before the sbp2 device
 	 * allows someone else to login instead. One second makes sense. */
-	set_current_state(TASK_INTERRUPTIBLE);
-	schedule_timeout(HZ);
+	ssleep(1);

Maybe the current code is _deliberately_ accepting interruption by signals but trying to complete sbp2_probe() anyway. However it seems more plausible to me to abort the device probe, for example like this:

Quoted 4 lines
if (msleep_interruptible(1000)) {
	sbp2_remove_device(scsi_id);
	return -EINTR;
}
Quoted 1 line

You might be right, but I'd like to get Ben's input on this, as I honeslty am

Don't hold your breath waiting for Ben's input. However, I would like to get one of the two proposed committed and tested by more users as this is a sore spot. I am not in a position at this time to fully research and test to make a call.

Quoted 8 lines

unsure. To be fair, I am trying to audit all usage of schedule_timeout() and the semantic interpretation (to me) of using TASK_INTERRUPTIBLE is that you wish to sleep a certain amount of time, but also are prepared for an early return on either signals or wait-queue events. msleep_interruptible() cleanly removes this second issue, but still requires the caller to respond appropriately if there is a return value. Hence, I like your change. I think it makes the most sense. Since I didn't/don't know how the device works, I was not able to make the change myself. Thanks for your input!

Sounds like a sign-off. Any other input before I request Stefan to make the final decision?

Quoted 5 lines
Quoted 1 line

Anyway, signal handling does not appear to be critical there.

Just out of curiousity, doesn't that run the risk, though, of signal_pending(current) being true for quite a bit of time following the timeout?

How much of this is "curiosity" vs a real risk?

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

#9 lore raw

Re: [UPDATE PATCH] ieee1394/sbp2: use ssleep() instead of schedule_timeout()

Quoted 18 lines · Dan Dennedy
Quoted 17 lines · Nishanth Aravamudan
Quoted 16 lines · Stefan Richter

drivers/ieee1394/sbp2.c alloc_fail: · line 905

Quoted 7 lines · Nishanth Aravamudan
drivers/ieee1394/sbp2.c +1 -2
@@ -902,8 +902,7 @@ alloc_fail:
 	 * connected to the sbp2 device being removed. That host would
 	 * have a certain amount of time to relogin before the sbp2 device
 	 * allows someone else to login instead. One second makes sense. */
-	set_current_state(TASK_INTERRUPTIBLE);
-	schedule_timeout(HZ);
+	ssleep(1);

Maybe the current code is _deliberately_ accepting interruption by signals but trying to complete sbp2_probe() anyway. However it seems more plausible to me to abort the device probe, for example like this:

if (msleep_interruptible(1000)) {
	sbp2_remove_device(scsi_id);
	return -EINTR;
}

[...]

Quoted 7 lines
Quoted 4 lines

I am trying to audit all usage of schedule_timeout() and the semantic interpretation (to me) of using TASK_INTERRUPTIBLE is that you wish to sleep a certain amount of time, but also are prepared for an early return on either signals or wait-queue events. [...]

Sounds like a sign-off. Any other input before I request Stefan to make the final decision?

Don't count on me here. I do not even know /which/ situations might introduce signals or wait queue events at this point. The only one that occurred to me is when nodemgr is about to be killed. In this situation, it is hardly beneficial to continue with the login. But there may be other events I do not know about which should not result in sbp2 giving up. Sorry, I should have been clear about this in my previous post.

Quoted 1 line
Quoted 1 line
Quoted 1 line

Anyway, signal handling does not appear to be critical there.

Or rather, it is not that important (although desirable) to always wait for 1000ms. This is just necessary for when another initiator was logged in into the target but did not reconnect or login again immediately after a bus reset. (Assuming the other initiator or the local host or the target require exclusive login, which is more common than concurrent login.)

Quoted 5 lines
Quoted 3 lines

Just out of curiousity, doesn't that run the risk, though, of signal_pending(current) being true for quite a bit of time following the timeout?

How much of this is "curiosity" vs a real risk?

Well, what might those events be? May we hold them off for one second? (Or perhaps even longer if we are about to login to several targets.) Should sbp2 proceeed to login when such events occur? I can't answer this for sure. Sorry,

Signature · 3 lines
Stefan Richter
-=====-=-=-= ---= -===-
http://arcgraph.de/sr/

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors

#10 lore raw

Re: [KJ] Re: [UPDATE PATCH] ieee1394/sbp2: use ssleep() instead of schedule_timeout()

In reply to Dan Dennedy
Quoted 45 lines · Dan Dennedy
Quoted 28 lines · Nishanth Aravamudan
Quoted 22 lines · Stefan Richter
Quoted 11 lines · Nishanth Aravamudan

Description: Use ssleep() instead of schedule_timeout() to guarantee the task
delays as expected. The existing code should not really need to run in TASK_INTERRUPTIBLE, as there is no check for signals (or even an early return
value whatsoever). ssleep() takes care of these issues.

--- 2.6.10-v/drivers/ieee1394/sbp2.c 2004-12-24 13:34:00.000000000
-0800
+++ 2.6.10/drivers/ieee1394/sbp2.c 2005-01-05 14:23:05.000000000 -0800

drivers/ieee1394/sbp2.c alloc_fail: · line 905

Quoted 6 lines
drivers/ieee1394/sbp2.c +1 -2
@@ -902,8 +902,7 @@ alloc_fail:
   * connected to the sbp2 device being removed. That host would
   * have a certain amount of time to relogin before the sbp2 device
   * allows someone else to login instead. One second makes sense. */
-  set_current_state(TASK_INTERRUPTIBLE);
-  schedule_timeout(HZ);
+  ssleep(1);

Maybe the current code is _deliberately_ accepting interruption by signals but trying to complete sbp2_probe() anyway. However it seems more plausible to me to abort the device probe, for example like this:

Quoted 4 lines
if (msleep_interruptible(1000)) {
    sbp2_remove_device(scsi_id);
    return -EINTR;
}
Quoted 1 line

You might be right, but I'd like to get Ben's input on this, as I honeslty am

Don't hold your breath waiting for Ben's input. However, I would like to get one of the two proposed committed and tested by more users as this is a sore spot. I am not in a position at this time to fully research and test to make a call.

Quoted 8 lines

unsure. To be fair, I am trying to audit all usage of schedule_timeout() and the semantic interpretation (to me) of using TASK_INTERRUPTIBLE is that you wish to sleep a certain amount of time, but also are prepared for an early return on either signals or wait-queue events. msleep_interruptible() cleanly removes this second issue, but still requires the caller to respond appropriately if there is a return value. Hence, I like your change. I think it makes the most sense. Since I didn't/don't know how the device works, I was not able to make the change myself. Thanks for your input!

Quoted 2 lines

Sounds like a sign-off. Any other input before I request Stefan to make the final decision?

Yes, this is an ACK for Stefan's change. Although the exact code he produced is not quite accurate. It would be most accurate to use

msleep_interruptible(1000);

if (signals_pending(current) {
       sbp2_remove_device(scsi_id);
       return -EINTR;
}

This accounts for the corner case when the sleep times out and a signal comes between the second-to-last and last jiffies. Thanks for both of your input! If you'd prefere me sending a new patch I can do so from work tomorrow.

Quoted 7 lines
Quoted 5 lines
Quoted 1 line

Anyway, signal handling does not appear to be critical there.

Just out of curiousity, doesn't that run the risk, though, of signal_pending(current) being true for quite a bit of time following the timeout?

How much of this is "curiosity" vs a real risk?

I think it should be ok, actually, the -EINTR should get passed back to userspace, where it would be handled appropriately. I hope :)

-Nish

_______________________________________________
Kernel-janitors mailing list
[email protected]
http://lists.osdl.org/mailman/listinfo/kernel-janitors