Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 20 Jan 2003 12:44:55 -0800 (PST)
From:      "Justin T. Gibbs" <gibbs@FreeBSD.org>
To:        cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org
Subject:   cvs commit: src/sys/dev/aic7xxx aic7xxx.c aic7xxx.h aic7xxx.reg aic7xxx.seq aic7xxx_inline.h aic7xxx_pci.c
Message-ID:  <200301202044.h0KKitb2033412@repoman.freebsd.org>

next in thread | raw e-mail | index | archive | help
gibbs       2003/01/20 12:44:55 PST

  Modified files:
    sys/dev/aic7xxx      aic7xxx.c aic7xxx.h aic7xxx.reg 
                         aic7xxx.seq aic7xxx_inline.h 
                         aic7xxx_pci.c 
  Log:
  aic7xxx.reg:
          Add a constant for the controller's stack size and the
          maximum scsi offset.
  
  aic7xxx.seq:
          Style nit.  The source is implied to be the destination
          unless overridden in an "and" instruction.
  
          Update target mode code for changes in identify seen
          sequencer flags.
  
  aic7xxx_pci.c:
          Ensure that the PCIERRGENDIS bit is set in the
          PCIERRGEN config space register.  Perhaps this
          is a reason for the spurios parity errors reported
          on U160 controllers.
  
          Honor the AHC_NO_BIOS_INIT flag.
  
          Allow PCI interrupt reporting to be disabled,
          by clearing the PERRRESEN bit in the command
          register.  This option is now enabled via a new
          softc flag: AHC_DISABLE_PCI_PERR.
  
          Disable SERR and pause the controller prior to performing
          our mmapped I/O test.  This should handle the case of
          controllers that do not "auto-access pause".  For legacy
          controllers, use SCB ram instead of scratch ram since
          the latter may contain settings left over from the BIOS
          that we will use if an seeprom is not found.
  
          Make use of new ahc_inl/outl() inlines.
  
  aic7xxx.h:
          Reformat a few comments to follow driver style.
  
          Add a controller flags that indicate that a controller
          has not been initialized by the BIOS and whether to
          disable PCI parity errors..
  
          Remove stack probing softc members.
  
          Add a few more syncrate constants that are useful in speed
          fallback calculations.
  
          Add the SHOW_MASKED_ERRORS debug flag.
  
  aic7xxx.h:
  aic7xxx.c:
          Implement the SCB_SILENT flag.  This is useful for
          hushing up the driver during DV or other operations
          that we expect to cause transmission errors.  The
          messages will still print if the SHOW_MASKED_ERRORS
          debug option is enabled.
  
  aic7xxx_inline.h:
          Implement ahc_[in|out][w|l|q].  This removes the need
          for manual 'or and shift" type operations throughout
          the driver.
  
  aic7xxx.c:
          Move SELTO dignostic so that the SCB is still valid
          when we use it for printing path information.
  
          If we are narrow, limit syncrate to Ultra2.
  
          Don't clobber ppr_options when forcing a renegotiation.
          The current ppr_options may be referenced while queuing
          new commands.  Don't set our width to unknown when forcing
          negotiation on narrow controllers.  This will confuse the
          negotiation code into negotiating with a wide message on
          narrow controllers.
  
          Add an "asserting atn" diagnostic with controller/target
          information.
  
          Remove the probe_stack code.  The stack is always
          4 deep on legacy controllers, so probing is pointless.
          This also avoids an issue where probing the stack would
          upset the aic7770.
  
          In ahc_reset(), record whether or not we found the
          controller in a reset state.  If the controller was
          already reset, assume that no BIOS has initialized
          the controller and ignore left over scratch ram
          settings.
  
          Fix an ifdef bug that caused sequencer debugging to
          be enabled always.
  
          Clear the ultraenb flag in our tstate during startup.
          The ultraenbled'ness of a device is recorded in the user
          transfer settings.  tstate->ultraenb bitmask indicates
          which devices we have negotiated an ultra speed with.
          Just after initialization, we are async.  Setting the
          ultraenb flag while async seems to be harmless, but it
          was confusing to see the ULTRAENB flag set in the SCB.
  
          Enhance residual diagnostic to indicate if the residual
          if for sense information or normal data transfers.
  
          Indicate the features, bugs, and flags set in the softc
          that are used to control firmware patch download when
          booting verbose.
  
          In ahc_dump_card_state() fix a logic reversal.  The
          SCSIPHASE register only exists on U160 controllers.
          The SCSISIGI register exists on all controllers.  Not
          the other way around.  Also print out the ERROR register.
  
          Allow ahc_dump_card_state() to be called when the sequencer
          is not paused.  Add dump card state markers as in the U320
          driver.
  
  Revision  Changes    Path
  1.88      +97 -94    src/sys/dev/aic7xxx/aic7xxx.c
  1.44      +32 -14    src/sys/dev/aic7xxx/aic7xxx.h
  1.43      +4 -1      src/sys/dev/aic7xxx/aic7xxx.reg
  1.123     +5 -4      src/sys/dev/aic7xxx/aic7xxx.seq
  1.20      +70 -1     src/sys/dev/aic7xxx/aic7xxx_inline.h
  1.22      +43 -24    src/sys/dev/aic7xxx/aic7xxx_pci.c

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200301202044.h0KKitb2033412>