aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAge
* sfc: Do not set net_device::trans_start in self-testBen Hutchings2009-11-26
| | | | | Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Move inline comment into kernel-docBen Hutchings2009-11-26
| | | | | Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Change MAC promiscuity and multicast hash at the same timeBen Hutchings2009-11-26
| | | | | | | | | | | | | | | | | From: Steve Hodgson <shodgson@solarflare.com> Currently we can set multicast hash immediately (in atomic context) but must delay setting MAC promiscuity. There is not that much point in deferring one but not the other, and setting the multicast hash on Siena will involve a firmware request. So process them both in efx_mac_work(). Also, set the broadcast bit in the multicast hash in efx_set_multicast_list(), since this is required for both Falcon and Siena. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Simplify XMAC link pollingBen Hutchings2009-11-26
| | | | | | | | | | | | | | | | | | | | | | | | From: Steve Hodgson <shodgson@solarflare.com> Only the XMAC on Falcon needs help from the driver to poll and reset the MAC-PHY link (XAUI); GMII is a simple parallel bus and on later NICs firmware takes care of the XAUI link. Also, an XMAC interrupt currently schedules a work item which simply clears a flag (efx_nic::mac_up) to be checked by the regular monitor (or the next link reconfiguration, if that is sooner). Rename the flag to xmac_poll_required, changing its sense. Remove the needless indirection and just set the flag immediately. Call falcon_xmac_poll() directly where required. Add a new generic operation mac_op::check_fault to check the link outside of regular monitoring, as required during self-tests. (Note that this leaves us with an unused work item, but we will immediately have another use for it.) Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Move Falcon board/PHY/MAC monitoring code to falcon.cBen Hutchings2009-11-26
| | | | | Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Split MAC stats DMA initiation and completionBen Hutchings2009-11-26
| | | | | | | | | | | | | | | From: Steve Hodgson <shodgson@solarflare.com> Currently we initiate MAC stats DMA and busy-wait for completion when stats are requested. We can improve on this with a periodic timer to initiate and poll for stats, and opportunistically poll when stats are requested. Since efx_nic::stats_disable_count and efx_stats_{disable,enable}() are Falcon-specific, rename them and move them accordingly. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Hold MAC lock for longer in efx_init_port()Ben Hutchings2009-11-26
| | | | | | | | | Although efx_init_port() is only called at probe time and so cannot race with port reconfiguration, most of the functions it calls can expect to be called with the MAC lock held. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Only switch Falcon MAC clocks as necessarySteve Hodgson2009-11-26
| | | | | Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Remove unused function efx_flush_queues()Ben Hutchings2009-11-26
| | | | | Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Fix bugs in RX queue flushingBen Hutchings2009-11-26
| | | | | | | | | Avoid overrunning the hardware limit of 4 concurrent RX queue flushes. Expand the queue flush state to support this. Make similar changes to TX flushing to keep the code symmetric. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Clean up struct falcon_board and struct falcon_board_dataBen Hutchings2009-11-26
| | | | | | | | Put all static information in struct falcon_board_type and replace it with a pointer in struct falcon_board. Simplify probing aocordingly. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Remove unused constantBen Hutchings2009-11-26
| | | | | Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Comment correctionsBen Hutchings2009-11-26
| | | | | Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Strengthen EFX_ASSERT_RESET_SERIALISEDBen Hutchings2009-11-26
| | | | | Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Treat all MAC registers as 128-bitBen Hutchings2009-11-26
| | | | | | | | | | Although all the defined fields in these registers are within 32 bits, they are architecturally defined as 128-bit like most other Falcon registers. In particular, we must use efx_reado() to ensure proper locking when reading MD_STAT_REG. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* sfc: Fix descriptor cache sizesBen Hutchings2009-11-26
| | | | | | | | These were accidentally undersized by a factor of 2, which limited performance. Signed-off-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* macvlan: export macvlan mode through netlinkArnd Bergmann2009-11-26
| | | | | | | | | | | | | | In order to support all three modes of macvlan at runtime, extend the existing netlink protocol to allow choosing the mode per macvlan slave interface. This depends on a matching patch to iproute2 in order to become accessible in user land. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* macvlan: implement bridge, VEPA and private modeArnd Bergmann2009-11-26
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This allows each macvlan slave device to be in one of three modes, depending on the use case: MACVLAN_PRIVATE: The device never communicates with any other device on the same upper_dev. This even includes frames coming back from a reflective relay, where supported by the adjacent bridge. MACVLAN_VEPA: The new Virtual Ethernet Port Aggregator (VEPA) mode, we assume that the adjacent bridge returns all frames where both source and destination are local to the macvlan port, i.e. the bridge is set up as a reflective relay. Broadcast frames coming in from the upper_dev get flooded to all macvlan interfaces in VEPA mode. We never deliver any frames locally. MACVLAN_BRIDGE: We provide the behavior of a simple bridge between different macvlan interfaces on the same port. Frames from one interface to another one get delivered directly and are not sent out externally. Broadcast frames get flooded to all other bridge ports and to the external interface, but when they come back from a reflective relay, we don't deliver them again. Since we know all the MAC addresses, the macvlan bridge mode does not require learning or STP like the bridge module does. Based on an earlier patch "macvlan: Reflect macvlan packets meant for other macvlan devices" by Eric Biederman. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Patrick McHardy <kaber@trash.net> Cc: Eric Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* macvlan: cleanup rx statisticsArnd Bergmann2009-11-26
| | | | | | | | | | | | | We have very similar code for rx statistics in two places in the macvlan driver, with a third one being added in the next patch. Consolidate them into one function to improve overall readability of the driver. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* veth: move loopback logic to common locationArnd Bergmann2009-11-26
| | | | | | | | | | | | | | | The veth driver contains code to forward an skb from the start_xmit function of one network device into the receive path of another device. Moving that code into a common location lets us reuse the code for direct forwarding of data between macvlan ports, and possibly in other drivers. Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* via-velocity: Bump versionSimon Kagstrom2009-11-26
| | | | | Signed-off-by: Simon Kagstrom <simon.kagstrom@netinsight.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* via-velocity: Set tx checksum from ethtool instead of module parameterSimon Kagstrom2009-11-26
| | | | | | | Defaults to on (as before). Signed-off-by: Simon Kagstrom <simon.kagstrom@netinsight.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* via-velocity: Re-enable transmit scatter-gather supportSimon Kagstrom2009-11-26
| | | | | | | | | | | | The velocity hardware can handle up to 7 memory segments. This can be turned on and off via ethtool. The support was removed in commit 83c98a8cd04dd0f848574370594886ba3bf56750 but is re-enabled and cleaned up here. It's off by default. Signed-off-by: Simon Kagstrom <simon.kagstrom@netinsight.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* via-velocity: Change DMA_LENGTH_DEF (from the VIA driver)Simon Kagstrom2009-11-26
| | | | | | | | | The VIA driver has changed the default for the DMA_LENGTH_DEF parameter. Together with adaptive interrupt supression and NAPI support, this improves performance quite a bit Signed-off-by: Simon Kagstrom <simon.kagstrom@netinsight.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* via-velocity: Implement NAPI supportSimon Kagstrom2009-11-26
| | | | | | | | | | This patch adds NAPI support for VIA velocity. The new velocity_poll function also pairs tx/rx handling twice which improves perforamance on some workloads (e.g., netperf UDP_STREAM) significantly (that part is from the VIA driver). Signed-off-by: Simon Kagstrom <simon.kagstrom@netinsight.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* via-velocity: Add ethtool interrupt coalescing supportSimon Kagstrom2009-11-26
| | | | | | | | | (Partially from the upstream VIA driver). Tweaking the number of frames-per-interrupt and timer-until-interrupt can reduce the amount of CPU work quite a lot. Signed-off-by: Simon Kagstrom <simon.kagstrom@netinsight.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* via-velocity: Correct 64-byte alignment for rx buffersSimon Kagstrom2009-11-26
| | | | | | | | | | | | | | | | | | | | | (From the VIA driver). The current code does not guarantee 64-byte alignment since it simply does int add = skb->data & 63; skb->data += add; (via skb_reserve). So for example, if the skb->data address would be 0x10, this would result in 32-byte alignment (0x10 + 0x10). Correct by adding 64 - (skb->data & 63) instead. Signed-off-by: Simon Kagstrom <simon.kagstrom@netinsight.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* r8169: move PHY regs tables to .rodataAlexey Dobriyan2009-11-25
| | | | | | | | | | | As side effect, consume less stack. -rtl8169_get_mac_version [vmlinux]: 432 -rtl8169_init_one [vmlinux]: 376 +rtl8169_init_one [vmlinux]: 136 Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* xfrm: Add SHA384 and SHA512 HMAC authentication algorithms to XFRMMartin Willi2009-11-25
| | | | | | | | | These algorithms use a truncation of 192/256 bits, as specified in RFC4868. Signed-off-by: Martin Willi <martin@strongswan.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
* xfrm: Use the user specified truncation length in ESP and AHMartin Willi2009-11-25
| | | | | | | | Instead of using the hardcoded truncation for authentication algorithms, use the truncation length specified on xfrm_state. Signed-off-by: Martin Willi <martin@strongswan.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* xfrm: Store aalg in xfrm_state with a user specified truncation lengthMartin Willi2009-11-25
| | | | | | | | | | | | | | Adding a xfrm_state requires an authentication algorithm specified either as xfrm_algo or as xfrm_algo_auth with a specific truncation length. For compatibility, both attributes are dumped to userspace, and we also accept both attributes, but prefer the new syntax. If no truncation length is specified, or the authentication algorithm is specified using xfrm_algo, the truncation length from the algorithm description in the kernel is used. Signed-off-by: Martin Willi <martin@strongswan.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* xfrm: Define new XFRM netlink auth attribute with specified truncation bitsMartin Willi2009-11-25
| | | | | | | | | | The new XFRMA_ALG_AUTH_TRUNC attribute taking a xfrm_algo_auth as argument allows the installation of authentication algorithms with a truncation length specified in userspace, i.e. SHA256 with 128 bit instead of 96 bit truncation. Signed-off-by: Martin Willi <martin@strongswan.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: convert /proc/net/rt_acct to seq_fileAlexey Dobriyan2009-11-25
| | | | | | | | | | | | Rewrite statistics accumulation to be in terms of structure fields, not raw u32 additions. Keep them in same order, though. This is the last user of create_proc_read_entry() in net/, please NAK all new ones as well as all new ->write_proc, ->read_proc and create_proc_entry() users. Cc me if there are problems. :-) Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: use net_eq to compare netsOctavian Purdila2009-11-25
| | | | | | | | | | | | | | | | | | | | | | | Generated with the following semantic patch @@ struct net *n1; struct net *n2; @@ - n1 == n2 + net_eq(n1, n2) @@ struct net *n1; struct net *n2; @@ - n1 != n2 + !net_eq(n1, n2) applied over {include,net,drivers/net}. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge branch 'master' of ↵David S. Miller2009-11-24
|\ | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6
| * ath9k: set ps_default as falseJohn W. Linville2009-11-23
| | | | | | | | | | | | | | | | | | | | | | Copied from original one-line patch here: http://bugzilla.kernel.org/show_bug.cgi?id=14267#c26 (This is for 2.6.33 and beyond, where the bool was changed to a flag by "cfg80211: convert bools into flags". -- JWL) Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ssb: Fix range check in sprom writeMichael Buesch2009-11-23
| | | | | | | | | | | | | | | | | | | | | | | | The range check in the sprom image parser hex2sprom() is broken. One sprom word is 4 hex characters. This fixes the check and also adds much better sanity checks to the code. We better make sure the image is OK by doing some sanity checks to avoid bricking the device by accident. Signed-off-by: Michael Buesch <mb@bu3sch.de> Cc: stable@kernel.org Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ssb: Fix SPROM writingMichael Buesch2009-11-23
| | | | | | | | | | | | | | | | | | | | | | The SPROM writing routines were broken since we rewrote the suspend handling on wireless devices, because SPROM writing depended on suspend. This patch changes it and freezes devices with the driver remove(), probe() callbacks instead. This also simplifies the whole logics a lot. Signed-off-by: Michael Buesch <mb@bu3sch.de> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * rndis_wlan: disable stall workaroundJussi Kivilinna2009-11-23
| | | | | | | | | | | | | | | | | | Stall workaround doesn't work with bcm4320a devices like with bcm4320b. This workaround actually causes more stalls/device freeze on bcm4320a. Therefore disable stall workaround by default. Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * rndis_wlan: fix buffer overflow in rndis_query_oidJussi Kivilinna2009-11-23
| | | | | | | | | | | | | | | | | | | | | | rndis_query_oid overwrites *len which stores buffer size to return full size of received command and then uses *len with memcpy to fill buffer with command. Ofcourse memcpy should be done before replacing buffer size. Signed-off-by: Jussi Kivilinna <jussi.kivilinna@mbnet.fi> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: Fix missing kernel-doc notationJaswinder Singh Rajput2009-11-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix the following htmldocs warnings: Warning(net/mac80211/sta_info.h:322): No description found for parameter 'drv_unblock_wk' Warning(net/mac80211/sta_info.h:322): No description found for parameter 'drv_unblock_wk' Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Cc: Johannes Berg <johannes@sipsolutions.net> Cc: John W. Linville <linville@tuxdriver.com> Cc: Randy Dunlap <randy.dunlap@oracle.com> Cc: David Miller <davem@davemloft.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: remove reset led_tpt from iwl_leds_init()Wey-Yi Guy2009-11-23
| | | | | | | | | | | | | | | | | | | | | | | | | | Current blinking rate is calculated based on the difference between current tx/rx byte counts and priv->led_tpt. priv->led_tpt should not get reset in iwl_leds_init(), this function can be called by bring interface "up" or "down", or when uCode sysassert occurred. resetting the led_tpt parameter will introduce incorrect led blinking behavior. Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: change message for cmd queue full errorWey-Yi Guy2009-11-23
| | | | | | | | | | | | | | | | | | Change error message for command queue full Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Acked-by: Ben Cahill <ben.m.cahill@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: print limited number of event log when uCode errorWey-Yi Guy2009-11-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | To help iwlagn uCode debugging, event log will dump to syslog when driver detect uCode error occurred, but this only happen when compile with CONFIG_IWLWIFI_DEBUG and debug flag is enabled; which is not always the case. Also, there is another problem, if the flag is set, the entire event log buffer will be dump to syslog, it can flood the syslog and make it very difficult to debug the problem. Change the default to only dump last 20 entries of event log to syslog unless the following condition meets: 1. both compile with CONFIG_IWLWIFI_DEBUG and debug flag is enabled, and then dump the entire event buffer to syslog. 2. dump event log request from debugfs Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Acked-by: Ben Cahill <ben.m.cahill@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: Tell the ucode immediately when association state changesJay Sternberg2009-11-23
| | | | | | | | | | | | | | | | | | | | | | | | When we get a state change of associated or not, we need to tell the ucode via the RX_ON command using the filter flags. This will prevent the ucode from sending any packets when not associated, specifically not sending NULL QOS packets after a deauthentication which causes the AP to repeatedly send deauth's in some situations. Signed-off-by: Jay Sternberg <jay.e.sternberg@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: dump error log when uCode error occurredWey-Yi Guy2009-11-23
| | | | | | | | | | | | | | | | | | | | | | | | | | uCode error log contain information as to what the error was and where it occurred necessary to debug any uCode issues. Always log the information without special debug flag, this can help to capture the important information when error happened. Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Acked-by: Ben Cahill <ben.m.cahill@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: by default, dump entire sram data portionWey-Yi Guy2009-11-23
| | | | | | | | | | | | | | | | | | | | | | For "sram" debugfs file, if user did not specify the offset and length, dump the entire data portion of sram by default. Data portion is 0x800000 - 0x80ffff, but the actual data size is known to the driver from the ucode file. Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: update supported PCI_ID list for 5xx0 seriesWey-Yi Guy2009-11-23
| | | | | | | | | | | | | | | | | | | | | | | | Update the PCI_ID list for 5xx0 series. Remove all the PCI_IDs which never made into production or not longer in production. Also make sure the supported bands(a/b/g/n) match specified PCI_IDs Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: separate IO tracingJohannes Berg2009-11-23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Since IO tracing is usually not needed and generates a lot of data, separate it into its own trace system so that we can always enable iwlwifi:* and not have to worry about getting too much data. If IO tracing is then really needed we can enable iwlwifi_io:* in addition and get that data. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * iwlwifi: fix reserved2 field in iwl4965_addstaWey-Yi Guy2009-11-23
| | | | | | | | | | | | | | | | reserved2 field in "struct iwl4965_addsta_cmd" is __le16. Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by: Reinette Chatre <reinette.chatre@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
generated by cgit v1.2.2 (git 2.25.0) at 2025-09-07 17:44:07 -0400
ref='#n1977'>1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 
/* esp_scsi.c: ESP SCSI driver.
 *
 * Copyright (C) 2007 David S. Miller (davem@davemloft.net)
 */

#include <linux/kernel.h>
#include <linux/types.h>
#include <linux/slab.h>
#include <linux/delay.h>
#include <linux/list.h>
#include <linux/completion.h>
#include <linux/kallsyms.h>
#include <linux/module.h>
#include <linux/moduleparam.h>
#include <linux/init.h>
#include <linux/irqreturn.h>

#include <asm/irq.h>
#include <asm/io.h>
#include <asm/dma.h>

#include <scsi/scsi.h>
#include <scsi/scsi_host.h>
#include <scsi/scsi_cmnd.h>
#include <scsi/scsi_device.h>
#include <scsi/scsi_tcq.h>
#include <scsi/scsi_dbg.h>
#include <scsi/scsi_transport_spi.h>

#include "esp_scsi.h"

#define DRV_MODULE_NAME		"esp"
#define PFX DRV_MODULE_NAME	": "
#define DRV_VERSION		"2.000"
#define DRV_MODULE_RELDATE	"April 19, 2007"

/* SCSI bus reset settle time in seconds.  */
static int esp_bus_reset_settle = 3;

static u32 esp_debug;
#define ESP_DEBUG_INTR		0x00000001
#define ESP_DEBUG_SCSICMD	0x00000002
#define ESP_DEBUG_RESET		0x00000004
#define ESP_DEBUG_MSGIN		0x00000008
#define ESP_DEBUG_MSGOUT	0x00000010
#define ESP_DEBUG_CMDDONE	0x00000020
#define ESP_DEBUG_DISCONNECT	0x00000040
#define ESP_DEBUG_DATASTART	0x00000080
#define ESP_DEBUG_DATADONE	0x00000100
#define ESP_DEBUG_RECONNECT	0x00000200
#define ESP_DEBUG_AUTOSENSE	0x00000400

#define esp_log_intr(f, a...) \
do {	if (esp_debug & ESP_DEBUG_INTR) \
		printk(f, ## a); \
} while (0)

#define esp_log_reset(f, a...) \
do {	if (esp_debug & ESP_DEBUG_RESET) \
		printk(f, ## a); \
} while (0)

#define esp_log_msgin(f, a...) \
do {	if (esp_debug & ESP_DEBUG_MSGIN) \
		printk(f, ## a); \
} while (0)

#define esp_log_msgout(f, a...) \
do {	if (esp_debug & ESP_DEBUG_MSGOUT) \
		printk(f, ## a); \
} while (0)

#define esp_log_cmddone(f, a...) \
do {	if (esp_debug & ESP_DEBUG_CMDDONE) \
		printk(f, ## a); \
} while (0)

#define esp_log_disconnect(f, a...) \
do {	if (esp_debug & ESP_DEBUG_DISCONNECT) \
		printk(f, ## a); \
} while (0)

#define esp_log_datastart(f, a...) \
do {	if (esp_debug & ESP_DEBUG_DATASTART) \
		printk(f, ## a); \
} while (0)

#define esp_log_datadone(f, a...) \
do {	if (esp_debug & ESP_DEBUG_DATADONE) \
		printk(f, ## a); \
} while (0)

#define esp_log_reconnect(f, a...) \
do {	if (esp_debug & ESP_DEBUG_RECONNECT) \
		printk(f, ## a); \
} while (0)

#define esp_log_autosense(f, a...) \
do {	if (esp_debug & ESP_DEBUG_AUTOSENSE) \
		printk(f, ## a); \
} while (0)

#define esp_read8(REG)		esp->ops->esp_read8(esp, REG)
#define esp_write8(VAL,REG)	esp->ops->esp_write8(esp, VAL, REG)

static void esp_log_fill_regs(struct esp *esp,
			      struct esp_event_ent *p)
{
	p->sreg = esp->sreg;
	p->seqreg = esp->seqreg;
	p->sreg2 = esp->sreg2;
	p->ireg = esp->ireg;
	p->select_state = esp->select_state;
	p->event = esp->event;
}

void scsi_esp_cmd(struct esp *esp, u8 val)
{
	struct esp_event_ent *p;
	int idx = esp->esp_event_cur;

	p = &esp->esp_event_log[idx];
	p->type = ESP_EVENT_TYPE_CMD;
	p->val = val;
	esp_log_fill_regs(esp, p);

	esp->esp_event_cur = (idx + 1) & (ESP_EVENT_LOG_SZ - 1);

	esp_write8(val, ESP_CMD);
}
EXPORT_SYMBOL(scsi_esp_cmd);

static void esp_event(struct esp *esp, u8 val)
{
	struct esp_event_ent *p;
	int idx = esp->esp_event_cur;

	p = &esp->esp_event_log[idx];
	p->type = ESP_EVENT_TYPE_EVENT;
	p->val = val;
	esp_log_fill_regs(esp, p);

	esp->esp_event_cur = (idx + 1) & (ESP_EVENT_LOG_SZ - 1);

	esp->event = val;
}

static void esp_dump_cmd_log(struct esp *esp)
{
	int idx = esp->esp_event_cur;
	int stop = idx;

	printk(KERN_INFO PFX "esp%d: Dumping command log\n",
	       esp->host->unique_id);
	do {
		struct esp_event_ent *p = &esp->esp_event_log[idx];

		printk(KERN_INFO PFX "esp%d: ent[%d] %s ",
		       esp->host->unique_id, idx,
		       p->type == ESP_EVENT_TYPE_CMD ? "CMD" : "EVENT");

		printk("val[%02x] sreg[%02x] seqreg[%02x] "
		       "sreg2[%02x] ireg[%02x] ss[%02x] event[%02x]\n",
		       p->val, p->sreg, p->seqreg,
		       p->sreg2, p->ireg, p->select_state, p->event);

		idx = (idx + 1) & (ESP_EVENT_LOG_SZ - 1);
	} while (idx != stop);
}

static void esp_flush_fifo(struct esp *esp)
{
	scsi_esp_cmd(esp, ESP_CMD_FLUSH);
	if (esp->rev == ESP236) {
		int lim = 1000;

		while (esp_read8(ESP_FFLAGS) & ESP_FF_FBYTES) {
			if (--lim == 0) {
				printk(KERN_ALERT PFX "esp%d: ESP_FF_BYTES "
				       "will not clear!\n",
				       esp->host->unique_id);
				break;
			}
			udelay(1);
		}
	}
}

static void hme_read_fifo(struct esp *esp)
{
	int fcnt = esp_read8(ESP_FFLAGS) & ESP_FF_FBYTES;
	int idx = 0;

	while (fcnt--) {
		esp->fifo[idx++] = esp_read8(ESP_FDATA);
		esp->fifo[idx++] = esp_read8(ESP_FDATA);
	}
	if (esp->sreg2 & ESP_STAT2_F1BYTE) {
		esp_write8(0, ESP_FDATA);
		esp->fifo[idx++] = esp_read8(ESP_FDATA);
		scsi_esp_cmd(esp, ESP_CMD_FLUSH);
	}
	esp->fifo_cnt = idx;
}

static void esp_set_all_config3(struct esp *esp, u8 val)
{
	int i;

	for (i = 0; i < ESP_MAX_TARGET; i++)
		esp->target[i].esp_config3 = val;
}

/* Reset the ESP chip, _not_ the SCSI bus. */
static void esp_reset_esp(struct esp *esp)
{
	u8 family_code, version;

	/* Now reset the ESP chip */
	scsi_esp_cmd(esp, ESP_CMD_RC);
	scsi_esp_cmd(esp, ESP_CMD_NULL | ESP_CMD_DMA);
	if (esp->rev == FAST)
		esp_write8(ESP_CONFIG2_FENAB, ESP_CFG2);
	scsi_esp_cmd(esp, ESP_CMD_NULL | ESP_CMD_DMA);

	/* This is the only point at which it is reliable to read
	 * the ID-code for a fast ESP chip variants.
	 */
	esp->max_period = ((35 * esp->ccycle) / 1000);
	if (esp->rev == FAST) {
		version = esp_read8(ESP_UID);
		family_code = (version & 0xf8) >> 3;
		if (family_code == 0x02)
			esp->rev = FAS236;
		else if (family_code == 0x0a)
			esp->rev = FASHME; /* Version is usually '5'. */
		else
			esp->rev = FAS100A;
		esp->min_period = ((4 * esp->ccycle) / 1000);
	} else {
		esp->min_period = ((5 * esp->ccycle) / 1000);
	}
	esp->max_period = (esp->max_period + 3)>>2;
	esp->min_period = (esp->min_period + 3)>>2;

	esp_write8(esp->config1, ESP_CFG1);
	switch (esp->rev) {
	case ESP100:
		/* nothing to do */
		break;

	case ESP100A:
		esp_write8(esp->config2, ESP_CFG2);
		break;

	case ESP236:
		/* Slow 236 */
		esp_write8(esp->config2, ESP_CFG2);
		esp->prev_cfg3 = esp->target[0].esp_config3;
		esp_write8(esp->prev_cfg3, ESP_CFG3);
		break;

	case FASHME:
		esp->config2 |= (ESP_CONFIG2_HME32 | ESP_CONFIG2_HMEFENAB);
		/* fallthrough... */

	case FAS236:
		/* Fast 236 or HME */
		esp_write8(esp->config2, ESP_CFG2);
		if (esp->rev == FASHME) {
			u8 cfg3 = esp->target[0].esp_config3;

			cfg3 |= ESP_CONFIG3_FCLOCK | ESP_CONFIG3_OBPUSH;
			if (esp->scsi_id >= 8)
				cfg3 |= ESP_CONFIG3_IDBIT3;
			esp_set_all_config3(esp, cfg3);
		} else {
			u32 cfg3 = esp->target[0].esp_config3;

			cfg3 |= ESP_CONFIG3_FCLK;
			esp_set_all_config3(esp, cfg3);
		}
		esp->prev_cfg3 = esp->target[0].esp_config3;
		esp_write8(esp->prev_cfg3, ESP_CFG3);
		if (esp->rev == FASHME) {
			esp->radelay = 80;
		} else {
			if (esp->flags & ESP_FLAG_DIFFERENTIAL)
				esp->radelay = 0;
			else
				esp->radelay = 96;
		}
		break;

	case FAS100A:
		/* Fast 100a */
		esp_write8(esp->config2, ESP_CFG2);
		esp_set_all_config3(esp,
				    (esp->target[0].esp_config3 |
				     ESP_CONFIG3_FCLOCK));
		esp->prev_cfg3 = esp->target[0].esp_config3;
		esp_write8(esp->prev_cfg3, ESP_CFG3);
		esp->radelay = 32;
		break;

	default:
		break;
	}

	/* Reload the configuration registers */
	esp_write8(esp->cfact, ESP_CFACT);

	esp->prev_stp = 0;
	esp_write8(esp->prev_stp, ESP_STP);

	esp->prev_soff = 0;
	esp_write8(esp->prev_soff, ESP_SOFF);

	esp_write8(esp->neg_defp, ESP_TIMEO);

	/* Eat any bitrot in the chip */
	esp_read8(ESP_INTRPT);
	udelay(100);
}

static void esp_map_dma(struct esp *esp, struct scsi_cmnd *cmd)
{
	struct esp_cmd_priv *spriv = ESP_CMD_PRIV(cmd);
	struct scatterlist *sg = scsi_sglist(cmd);
	int dir = cmd->sc_data_direction;
	int total, i;

	if (dir == DMA_NONE)
		return;

	spriv->u.num_sg = esp->ops->map_sg(esp, sg, scsi_sg_count(cmd), dir);
	spriv->cur_residue = sg_dma_len(sg);
	spriv->cur_sg = sg;

	total = 0;
	for (i = 0; i < spriv->u.num_sg; i++)
		total += sg_dma_len(&sg[i]);
	spriv->tot_residue = total;
}

static dma_addr_t esp_cur_dma_addr(struct esp_cmd_entry *ent,
				   struct scsi_cmnd *cmd)
{
	struct esp_cmd_priv *p = ESP_CMD_PRIV(cmd);

	if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
		return ent->sense_dma +
			(ent->sense_ptr - cmd->sense_buffer);
	}

	return sg_dma_address(p->cur_sg) +
		(sg_dma_len(p->cur_sg) -
		 p->cur_residue);
}

static unsigned int esp_cur_dma_len(struct esp_cmd_entry *ent,
				    struct scsi_cmnd *cmd)
{
	struct esp_cmd_priv *p = ESP_CMD_PRIV(cmd);

	if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
		return SCSI_SENSE_BUFFERSIZE -
			(ent->sense_ptr - cmd->sense_buffer);
	}
	return p->cur_residue;
}

static void esp_advance_dma(struct esp *esp, struct esp_cmd_entry *ent,
			    struct scsi_cmnd *cmd, unsigned int len)
{
	struct esp_cmd_priv *p = ESP_CMD_PRIV(cmd);

	if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
		ent->sense_ptr += len;
		return;
	}

	p->cur_residue -= len;
	p->tot_residue -= len;
	if (p->cur_residue < 0 || p->tot_residue < 0) {
		printk(KERN_ERR PFX "esp%d: Data transfer overflow.\n",
		       esp->host->unique_id);
		printk(KERN_ERR PFX "esp%d: cur_residue[%d] tot_residue[%d] "
		       "len[%u]\n",
		       esp->host->unique_id,
		       p->cur_residue, p->tot_residue, len);
		p->cur_residue = 0;
		p->tot_residue = 0;
	}
	if (!p->cur_residue && p->tot_residue) {
		p->cur_sg++;
		p->cur_residue = sg_dma_len(p->cur_sg);
	}
}

static void esp_unmap_dma(struct esp *esp, struct scsi_cmnd *cmd)
{
	struct esp_cmd_priv *spriv = ESP_CMD_PRIV(cmd);
	int dir = cmd->sc_data_direction;

	if (dir == DMA_NONE)
		return;

	esp->ops->unmap_sg(esp, scsi_sglist(cmd), spriv->u.num_sg, dir);
}

static void esp_save_pointers(struct esp *esp, struct esp_cmd_entry *ent)
{
	struct scsi_cmnd *cmd = ent->cmd;
	struct esp_cmd_priv *spriv = ESP_CMD_PRIV(cmd);

	if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
		ent->saved_sense_ptr = ent->sense_ptr;
		return;
	}
	ent->saved_cur_residue = spriv->cur_residue;
	ent->saved_cur_sg = spriv->cur_sg;
	ent->saved_tot_residue = spriv->tot_residue;
}

static void esp_restore_pointers(struct esp *esp, struct esp_cmd_entry *ent)
{
	struct scsi_cmnd *cmd = ent->cmd;
	struct esp_cmd_priv *spriv = ESP_CMD_PRIV(cmd);

	if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
		ent->sense_ptr = ent->saved_sense_ptr;
		return;
	}
	spriv->cur_residue = ent->saved_cur_residue;
	spriv->cur_sg = ent->saved_cur_sg;
	spriv->tot_residue = ent->saved_tot_residue;
}

static void esp_check_command_len(struct esp *esp, struct scsi_cmnd *cmd)
{
	if (cmd->cmd_len == 6 ||
	    cmd->cmd_len == 10 ||
	    cmd->cmd_len == 12) {
		esp->flags &= ~ESP_FLAG_DOING_SLOWCMD;
	} else {
		esp->flags |= ESP_FLAG_DOING_SLOWCMD;
	}
}

static void esp_write_tgt_config3(struct esp *esp, int tgt)
{
	if (esp->rev > ESP100A) {
		u8 val = esp->target[tgt].esp_config3;

		if (val != esp->prev_cfg3) {
			esp->prev_cfg3 = val;
			esp_write8(val, ESP_CFG3);
		}
	}
}

static void esp_write_tgt_sync(struct esp *esp, int tgt)
{
	u8 off = esp->target[tgt].esp_offset;
	u8 per = esp->target[tgt].esp_period;

	if (off != esp->prev_soff) {
		esp->prev_soff = off;
		esp_write8(off, ESP_SOFF);
	}
	if (per != esp->prev_stp) {
		esp->prev_stp = per;
		esp_write8(per, ESP_STP);
	}
}

static u32 esp_dma_length_limit(struct esp *esp, u32 dma_addr, u32 dma_len)
{
	if (esp->rev == FASHME) {
		/* Arbitrary segment boundaries, 24-bit counts.  */
		if (dma_len > (1U << 24))
			dma_len = (1U << 24);
	} else {
		u32 base, end;

		/* ESP chip limits other variants by 16-bits of transfer
		 * count.  Actually on FAS100A and FAS236 we could get
		 * 24-bits of transfer count by enabling ESP_CONFIG2_FENAB
		 * in the ESP_CFG2 register but that causes other unwanted
		 * changes so we don't use it currently.
		 */
		if (dma_len > (1U << 16))
			dma_len = (1U << 16);

		/* All of the DMA variants hooked up to these chips
		 * cannot handle crossing a 24-bit address boundary.
		 */
		base = dma_addr & ((1U << 24) - 1U);
		end = base + dma_len;
		if (end > (1U << 24))
			end = (1U <<24);
		dma_len = end - base;
	}
	return dma_len;
}

static int esp_need_to_nego_wide(struct esp_target_data *tp)
{
	struct scsi_target *target = tp->starget;

	return spi_width(target) != tp->nego_goal_width;
}

static int esp_need_to_nego_sync(struct esp_target_data *tp)
{
	struct scsi_target *target = tp->starget;

	/* When offset is zero, period is "don't care".  */
	if (!spi_offset(target) && !tp->nego_goal_offset)
		return 0;

	if (spi_offset(target) == tp->nego_goal_offset &&
	    spi_period(target) == tp->nego_goal_period)
		return 0;

	return 1;
}

static int esp_alloc_lun_tag(struct esp_cmd_entry *ent,
			     struct esp_lun_data *lp)
{
	if (!ent->tag[0]) {
		/* Non-tagged, slot already taken?  */
		if (lp->non_tagged_cmd)
			return -EBUSY;

		if (lp->hold) {
			/* We are being held by active tagged
			 * commands.
			 */
			if (lp->num_tagged)
				return -EBUSY;

			/* Tagged commands completed, we can unplug
			 * the queue and run this untagged command.
			 */
			lp->hold = 0;
		} else if (lp->num_tagged) {
			/* Plug the queue until num_tagged decreases
			 * to zero in esp_free_lun_tag.
			 */
			lp->hold = 1;
			return -EBUSY;
		}

		lp->non_tagged_cmd = ent;
		return 0;
	} else {
		/* Tagged command, see if blocked by a
		 * non-tagged one.
		 */
		if (lp->non_tagged_cmd || lp->hold)
			return -EBUSY;
	}

	BUG_ON(lp->tagged_cmds[ent->tag[1]]);

	lp->tagged_cmds[ent->tag[1]] = ent;
	lp->num_tagged++;

	return 0;
}

static void esp_free_lun_tag(struct esp_cmd_entry *ent,
			     struct esp_lun_data *lp)
{
	if (ent->tag[0]) {
		BUG_ON(lp->tagged_cmds[ent->tag[1]] != ent);
		lp->tagged_cmds[ent->tag[1]] = NULL;
		lp->num_tagged--;
	} else {
		BUG_ON(lp->non_tagged_cmd != ent);
		lp->non_tagged_cmd = NULL;
	}
}

/* When a contingent allegiance conditon is created, we force feed a
 * REQUEST_SENSE command to the device to fetch the sense data.  I
 * tried many other schemes, relying on the scsi error handling layer
 * to send out the REQUEST_SENSE automatically, but this was difficult
 * to get right especially in the presence of applications like smartd
 * which use SG_IO to send out their own REQUEST_SENSE commands.
 */
static void esp_autosense(struct esp *esp, struct esp_cmd_entry *ent)
{
	struct scsi_cmnd *cmd = ent->cmd;
	struct scsi_device *dev = cmd->device;
	int tgt, lun;
	u8 *p, val;

	tgt = dev->id;
	lun = dev->lun;


	if (!ent->sense_ptr) {
		esp_log_autosense("esp%d: Doing auto-sense for "
				  "tgt[%d] lun[%d]\n",
				  esp->host->unique_id, tgt, lun);

		ent->sense_ptr = cmd->sense_buffer;
		ent->sense_dma = esp->ops->map_single(esp,
						      ent->sense_ptr,
						      SCSI_SENSE_BUFFERSIZE,
						      DMA_FROM_DEVICE);
	}
	ent->saved_sense_ptr = ent->sense_ptr;

	esp->active_cmd = ent;

	p = esp->command_block;
	esp->msg_out_len = 0;

	*p++ = IDENTIFY(0, lun);
	*p++ = REQUEST_SENSE;
	*p++ = ((dev->scsi_level <= SCSI_2) ?
		(lun << 5) : 0);
	*p++ = 0;
	*p++ = 0;
	*p++ = SCSI_SENSE_BUFFERSIZE;
	*p++ = 0;

	esp->select_state = ESP_SELECT_BASIC;

	val = tgt;
	if (esp->rev == FASHME)
		val |= ESP_BUSID_RESELID | ESP_BUSID_CTR32BIT;
	esp_write8(val, ESP_BUSID);

	esp_write_tgt_sync(esp, tgt);
	esp_write_tgt_config3(esp, tgt);

	val = (p - esp->command_block);

	if (esp->rev == FASHME)
		scsi_esp_cmd(esp, ESP_CMD_FLUSH);
	esp->ops->send_dma_cmd(esp, esp->command_block_dma,
			       val, 16, 0, ESP_CMD_DMA | ESP_CMD_SELA);
}

static struct esp_cmd_entry *find_and_prep_issuable_command(struct esp *esp)
{
	struct esp_cmd_entry *ent;

	list_for_each_entry(ent, &esp->queued_cmds, list) {
		struct scsi_cmnd *cmd = ent->cmd;
		struct scsi_device *dev = cmd->device;
		struct esp_lun_data *lp = dev->hostdata;

		if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
			ent->tag[0] = 0;
			ent->tag[1] = 0;
			return ent;
		}

		if (!scsi_populate_tag_msg(cmd, &ent->tag[0])) {
			ent->tag[0] = 0;
			ent->tag[1] = 0;
		}

		if (esp_alloc_lun_tag(ent, lp) < 0)
			continue;

		return ent;
	}

	return NULL;
}

static void esp_maybe_execute_command(struct esp *esp)
{
	struct esp_target_data *tp;
	struct esp_lun_data *lp;
	struct scsi_device *dev;
	struct scsi_cmnd *cmd;
	struct esp_cmd_entry *ent;
	int tgt, lun, i;
	u32 val, start_cmd;
	u8 *p;

	if (esp->active_cmd ||
	    (esp->flags & ESP_FLAG_RESETTING))
		return;

	ent = find_and_prep_issuable_command(esp);
	if (!ent)
		return;

	if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
		esp_autosense(esp, ent);
		return;
	}

	cmd = ent->cmd;
	dev = cmd->device;
	tgt = dev->id;
	lun = dev->lun;
	tp = &esp->target[tgt];
	lp = dev->hostdata;

	list_move(&ent->list, &esp->active_cmds);

	esp->active_cmd = ent;

	esp_map_dma(esp, cmd);
	esp_save_pointers(esp, ent);

	esp_check_command_len(esp, cmd);

	p = esp->command_block;

	esp->msg_out_len = 0;
	if (tp->flags & ESP_TGT_CHECK_NEGO) {
		/* Need to negotiate.  If the target is broken
		 * go for synchronous transfers and non-wide.
		 */
		if (tp->flags & ESP_TGT_BROKEN) {
			tp->flags &= ~ESP_TGT_DISCONNECT;
			tp->nego_goal_period = 0;
			tp->nego_goal_offset = 0;
			tp->nego_goal_width = 0;
			tp->nego_goal_tags = 0;
		}

		/* If the settings are not changing, skip this.  */
		if (spi_width(tp->starget) == tp->nego_goal_width &&
		    spi_period(tp->starget) == tp->nego_goal_period &&
		    spi_offset(tp->starget) == tp->nego_goal_offset) {
			tp->flags &= ~ESP_TGT_CHECK_NEGO;
			goto build_identify;
		}

		if (esp->rev == FASHME && esp_need_to_nego_wide(tp)) {
			esp->msg_out_len =
				spi_populate_width_msg(&esp->msg_out[0],
						       (tp->nego_goal_width ?
							1 : 0));
			tp->flags |= ESP_TGT_NEGO_WIDE;
		} else if (esp_need_to_nego_sync(tp)) {
			esp->msg_out_len =
				spi_populate_sync_msg(&esp->msg_out[0],
						      tp->nego_goal_period,
						      tp->nego_goal_offset);
			tp->flags |= ESP_TGT_NEGO_SYNC;
		} else {
			tp->flags &= ~ESP_TGT_CHECK_NEGO;
		}

		/* Process it like a slow command.  */
		if (tp->flags & (ESP_TGT_NEGO_WIDE | ESP_TGT_NEGO_SYNC))
			esp->flags |= ESP_FLAG_DOING_SLOWCMD;
	}

build_identify:
	/* If we don't have a lun-data struct yet, we're probing
	 * so do not disconnect.  Also, do not disconnect unless
	 * we have a tag on this command.
	 */
	if (lp && (tp->flags & ESP_TGT_DISCONNECT) && ent->tag[0])
		*p++ = IDENTIFY(1, lun);
	else
		*p++ = IDENTIFY(0, lun);

	if (ent->tag[0] && esp->rev == ESP100) {
		/* ESP100 lacks select w/atn3 command, use select
		 * and stop instead.
		 */
		esp->flags |= ESP_FLAG_DOING_SLOWCMD;
	}

	if (!(esp->flags & ESP_FLAG_DOING_SLOWCMD)) {
		start_cmd = ESP_CMD_DMA | ESP_CMD_SELA;
		if (ent->tag[0]) {
			*p++ = ent->tag[0];
			*p++ = ent->tag[1];

			start_cmd = ESP_CMD_DMA | ESP_CMD_SA3;
		}

		for (i = 0; i < cmd->cmd_len; i++)
			*p++ = cmd->cmnd[i];

		esp->select_state = ESP_SELECT_BASIC;
	} else {
		esp->cmd_bytes_left = cmd->cmd_len;
		esp->cmd_bytes_ptr = &cmd->cmnd[0];

		if (ent->tag[0]) {
			for (i = esp->msg_out_len - 1;
			     i >= 0; i--)
				esp->msg_out[i + 2] = esp->msg_out[i];
			esp->msg_out[0] = ent->tag[0];
			esp->msg_out[1] = ent->tag[1];
			esp->msg_out_len += 2;
		}

		start_cmd = ESP_CMD_DMA | ESP_CMD_SELAS;
		esp->select_state = ESP_SELECT_MSGOUT;
	}
	val = tgt;
	if (esp->rev == FASHME)
		val |= ESP_BUSID_RESELID | ESP_BUSID_CTR32BIT;
	esp_write8(val, ESP_BUSID);

	esp_write_tgt_sync(esp, tgt);
	esp_write_tgt_config3(esp, tgt);

	val = (p - esp->command_block);

	if (esp_debug & ESP_DEBUG_SCSICMD) {
		printk("ESP: tgt[%d] lun[%d] scsi_cmd [ ", tgt, lun);
		for (i = 0; i < cmd->cmd_len; i++)
			printk("%02x ", cmd->cmnd[i]);
		printk("]\n");
	}

	if (esp->rev == FASHME)
		scsi_esp_cmd(esp, ESP_CMD_FLUSH);
	esp->ops->send_dma_cmd(esp, esp->command_block_dma,
			       val, 16, 0, start_cmd);
}

static struct esp_cmd_entry *esp_get_ent(struct esp *esp)
{
	struct list_head *head = &esp->esp_cmd_pool;
	struct esp_cmd_entry *ret;

	if (list_empty(head)) {
		ret = kzalloc(sizeof(struct esp_cmd_entry), GFP_ATOMIC);
	} else {
		ret = list_entry(head->next, struct esp_cmd_entry, list);
		list_del(&ret->list);
		memset(ret, 0, sizeof(*ret));
	}
	return ret;
}

static void esp_put_ent(struct esp *esp, struct esp_cmd_entry *ent)
{
	list_add(&ent->list, &esp->esp_cmd_pool);
}

static void esp_cmd_is_done(struct esp *esp, struct esp_cmd_entry *ent,
			    struct scsi_cmnd *cmd, unsigned int result)
{
	struct scsi_device *dev = cmd->device;
	int tgt = dev->id;
	int lun = dev->lun;

	esp->active_cmd = NULL;
	esp_unmap_dma(esp, cmd);
	esp_free_lun_tag(ent, dev->hostdata);
	cmd->result = result;

	if (ent->eh_done) {
		complete(ent->eh_done);
		ent->eh_done = NULL;
	}

	if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
		esp->ops->unmap_single(esp, ent->sense_dma,
				       SCSI_SENSE_BUFFERSIZE, DMA_FROM_DEVICE);
		ent->sense_ptr = NULL;

		/* Restore the message/status bytes to what we actually
		 * saw originally.  Also, report that we are providing
		 * the sense data.
		 */
		cmd->result = ((DRIVER_SENSE << 24) |
			       (DID_OK << 16) |
			       (COMMAND_COMPLETE << 8) |
			       (SAM_STAT_CHECK_CONDITION << 0));

		ent->flags &= ~ESP_CMD_FLAG_AUTOSENSE;
		if (esp_debug & ESP_DEBUG_AUTOSENSE) {
			int i;

			printk("esp%d: tgt[%d] lun[%d] AUTO SENSE[ ",
			       esp->host->unique_id, tgt, lun);
			for (i = 0; i < 18; i++)
				printk("%02x ", cmd->sense_buffer[i]);
			printk("]\n");
		}
	}

	cmd->scsi_done(cmd);

	list_del(&ent->list);
	esp_put_ent(esp, ent);

	esp_maybe_execute_command(esp);
}

static unsigned int compose_result(unsigned int status, unsigned int message,
				   unsigned int driver_code)
{
	return (status | (message << 8) | (driver_code << 16));
}

static void esp_event_queue_full(struct esp *esp, struct esp_cmd_entry *ent)
{
	struct scsi_device *dev = ent->cmd->device;
	struct esp_lun_data *lp = dev->hostdata;

	scsi_track_queue_full(dev, lp->num_tagged - 1);
}

static int esp_queuecommand_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
{
	struct scsi_device *dev = cmd->device;
	struct esp *esp = shost_priv(dev->host);
	struct esp_cmd_priv *spriv;
	struct esp_cmd_entry *ent;

	ent = esp_get_ent(esp);
	if (!ent)
		return SCSI_MLQUEUE_HOST_BUSY;

	ent->cmd = cmd;

	cmd->scsi_done = done;

	spriv = ESP_CMD_PRIV(cmd);
	spriv->u.dma_addr = ~(dma_addr_t)0x0;

	list_add_tail(&ent->list, &esp->queued_cmds);

	esp_maybe_execute_command(esp);

	return 0;
}

static DEF_SCSI_QCMD(esp_queuecommand)

static int esp_check_gross_error(struct esp *esp)
{
	if (esp->sreg & ESP_STAT_SPAM) {
		/* Gross Error, could be one of:
		 * - top of fifo overwritten
		 * - top of command register overwritten
		 * - DMA programmed with wrong direction
		 * - improper phase change
		 */
		printk(KERN_ERR PFX "esp%d: Gross error sreg[%02x]\n",
		       esp->host->unique_id, esp->sreg);
		/* XXX Reset the chip. XXX */
		return 1;
	}
	return 0;
}

static int esp_check_spur_intr(struct esp *esp)
{
	switch (esp->rev) {
	case ESP100:
	case ESP100A:
		/* The interrupt pending bit of the status register cannot
		 * be trusted on these revisions.
		 */
		esp->sreg &= ~ESP_STAT_INTR;
		break;

	default:
		if (!(esp->sreg & ESP_STAT_INTR)) {
			esp->ireg = esp_read8(ESP_INTRPT);
			if (esp->ireg & ESP_INTR_SR)
				return 1;

			/* If the DMA is indicating interrupt pending and the
			 * ESP is not, the only possibility is a DMA error.
			 */
			if (!esp->ops->dma_error(esp)) {
				printk(KERN_ERR PFX "esp%d: Spurious irq, "
				       "sreg=%02x.\n",
				       esp->host->unique_id, esp->sreg);
				return -1;
			}

			printk(KERN_ERR PFX "esp%d: DMA error\n",
			       esp->host->unique_id);

			/* XXX Reset the chip. XXX */
			return -1;
		}
		break;
	}

	return 0;
}

static void esp_schedule_reset(struct esp *esp)
{
	esp_log_reset("ESP: esp_schedule_reset() from %pf\n",
		      __builtin_return_address(0));
	esp->flags |= ESP_FLAG_RESETTING;
	esp_event(esp, ESP_EVENT_RESET);
}

/* In order to avoid having to add a special half-reconnected state
 * into the driver we just sit here and poll through the rest of
 * the reselection process to get the tag message bytes.
 */
static struct esp_cmd_entry *esp_reconnect_with_tag(struct esp *esp,
						    struct esp_lun_data *lp)
{
	struct esp_cmd_entry *ent;
	int i;

	if (!lp->num_tagged) {
		printk(KERN_ERR PFX "esp%d: Reconnect w/num_tagged==0\n",
		       esp->host->unique_id);
		return NULL;
	}

	esp_log_reconnect("ESP: reconnect tag, ");

	for (i = 0; i < ESP_QUICKIRQ_LIMIT; i++) {
		if (esp->ops->irq_pending(esp))
			break;
	}
	if (i == ESP_QUICKIRQ_LIMIT) {
		printk(KERN_ERR PFX "esp%d: Reconnect IRQ1 timeout\n",
		       esp->host->unique_id);
		return NULL;
	}

	esp->sreg = esp_read8(ESP_STATUS);
	esp->ireg = esp_read8(ESP_INTRPT);

	esp_log_reconnect("IRQ(%d:%x:%x), ",
			  i, esp->ireg, esp->sreg);

	if (esp->ireg & ESP_INTR_DC) {
		printk(KERN_ERR PFX "esp%d: Reconnect, got disconnect.\n",
		       esp->host->unique_id);
		return NULL;
	}

	if ((esp->sreg & ESP_STAT_PMASK) != ESP_MIP) {
		printk(KERN_ERR PFX "esp%d: Reconnect, not MIP sreg[%02x].\n",
		       esp->host->unique_id, esp->sreg);
		return NULL;
	}

	/* DMA in the tag bytes... */
	esp->command_block[0] = 0xff;
	esp->command_block[1] = 0xff;
	esp->ops->send_dma_cmd(esp, esp->command_block_dma,
			       2, 2, 1, ESP_CMD_DMA | ESP_CMD_TI);

	/* ACK the message.  */
	scsi_esp_cmd(esp, ESP_CMD_MOK);

	for (i = 0; i < ESP_RESELECT_TAG_LIMIT; i++) {
		if (esp->ops->irq_pending(esp)) {
			esp->sreg = esp_read8(ESP_STATUS);
			esp->ireg = esp_read8(ESP_INTRPT);
			if (esp->ireg & ESP_INTR_FDONE)
				break;
		}
		udelay(1);
	}
	if (i == ESP_RESELECT_TAG_LIMIT) {
		printk(KERN_ERR PFX "esp%d: Reconnect IRQ2 timeout\n",
		       esp->host->unique_id);
		return NULL;
	}
	esp->ops->dma_drain(esp);
	esp->ops->dma_invalidate(esp);

	esp_log_reconnect("IRQ2(%d:%x:%x) tag[%x:%x]\n",
			  i, esp->ireg, esp->sreg,
			  esp->command_block[0],
			  esp->command_block[1]);

	if (esp->command_block[0] < SIMPLE_QUEUE_TAG ||
	    esp->command_block[0] > ORDERED_QUEUE_TAG) {
		printk(KERN_ERR PFX "esp%d: Reconnect, bad tag "
		       "type %02x.\n",
		       esp->host->unique_id, esp->command_block[0]);
		return NULL;
	}

	ent = lp->tagged_cmds[esp->command_block[1]];
	if (!ent) {
		printk(KERN_ERR PFX "esp%d: Reconnect, no entry for "
		       "tag %02x.\n",
		       esp->host->unique_id, esp->command_block[1]);
		return NULL;
	}

	return ent;
}

static int esp_reconnect(struct esp *esp)
{
	struct esp_cmd_entry *ent;
	struct esp_target_data *tp;
	struct esp_lun_data *lp;
	struct scsi_device *dev;
	int target, lun;

	BUG_ON(esp->active_cmd);
	if (esp->rev == FASHME) {
		/* FASHME puts the target and lun numbers directly
		 * into the fifo.
		 */
		target = esp->fifo[0];
		lun = esp->fifo[1] & 0x7;
	} else {
		u8 bits = esp_read8(ESP_FDATA);

		/* Older chips put the lun directly into the fifo, but
		 * the target is given as a sample of the arbitration
		 * lines on the bus at reselection time.  So we should
		 * see the ID of the ESP and the one reconnecting target
		 * set in the bitmap.
		 */
		if (!(bits & esp->scsi_id_mask))
			goto do_reset;
		bits &= ~esp->scsi_id_mask;
		if (!bits || (bits & (bits - 1)))
			goto do_reset;

		target = ffs(bits) - 1;
		lun = (esp_read8(ESP_FDATA) & 0x7);

		scsi_esp_cmd(esp, ESP_CMD_FLUSH);
		if (esp->rev == ESP100) {
			u8 ireg = esp_read8(ESP_INTRPT);
			/* This chip has a bug during reselection that can
			 * cause a spurious illegal-command interrupt, which
			 * we simply ACK here.  Another possibility is a bus
			 * reset so we must check for that.
			 */
			if (ireg & ESP_INTR_SR)
				goto do_reset;
		}
		scsi_esp_cmd(esp, ESP_CMD_NULL);
	}

	esp_write_tgt_sync(esp, target);
	esp_write_tgt_config3(esp, target);

	scsi_esp_cmd(esp, ESP_CMD_MOK);

	if (esp->rev == FASHME)
		esp_write8(target | ESP_BUSID_RESELID | ESP_BUSID_CTR32BIT,
			   ESP_BUSID);

	tp = &esp->target[target];
	dev = __scsi_device_lookup_by_target(tp->starget, lun);
	if (!dev) {
		printk(KERN_ERR PFX "esp%d: Reconnect, no lp "
		       "tgt[%u] lun[%u]\n",
		       esp->host->unique_id, target, lun);
		goto do_reset;
	}
	lp = dev->hostdata;

	ent = lp->non_tagged_cmd;
	if (!ent) {
		ent = esp_reconnect_with_tag(esp, lp);
		if (!ent)
			goto do_reset;
	}

	esp->active_cmd = ent;

	if (ent->flags & ESP_CMD_FLAG_ABORT) {
		esp->msg_out[0] = ABORT_TASK_SET;
		esp->msg_out_len = 1;
		scsi_esp_cmd(esp, ESP_CMD_SATN);
	}

	esp_event(esp, ESP_EVENT_CHECK_PHASE);
	esp_restore_pointers(esp, ent);
	esp->flags |= ESP_FLAG_QUICKIRQ_CHECK;
	return 1;

do_reset:
	esp_schedule_reset(esp);
	return 0;
}

static int esp_finish_select(struct esp *esp)
{
	struct esp_cmd_entry *ent;
	struct scsi_cmnd *cmd;
	u8 orig_select_state;

	orig_select_state = esp->select_state;

	/* No longer selecting.  */
	esp->select_state = ESP_SELECT_NONE;

	esp->seqreg = esp_read8(ESP_SSTEP) & ESP_STEP_VBITS;
	ent = esp->active_cmd;
	cmd = ent->cmd;

	if (esp->ops->dma_error(esp)) {
		/* If we see a DMA error during or as a result of selection,
		 * all bets are off.
		 */
		esp_schedule_reset(esp);
		esp_cmd_is_done(esp, ent, cmd, (DID_ERROR << 16));
		return 0;
	}

	esp->ops->dma_invalidate(esp);

	if (esp->ireg == (ESP_INTR_RSEL | ESP_INTR_FDONE)) {
		struct esp_target_data *tp = &esp->target[cmd->device->id];

		/* Carefully back out of the selection attempt.  Release
		 * resources (such as DMA mapping & TAG) and reset state (such
		 * as message out and command delivery variables).
		 */
		if (!(ent->flags & ESP_CMD_FLAG_AUTOSENSE)) {
			esp_unmap_dma(esp, cmd);
			esp_free_lun_tag(ent, cmd->device->hostdata);
			tp->flags &= ~(ESP_TGT_NEGO_SYNC | ESP_TGT_NEGO_WIDE);
			esp->flags &= ~ESP_FLAG_DOING_SLOWCMD;
			esp->cmd_bytes_ptr = NULL;
			esp->cmd_bytes_left = 0;
		} else {
			esp->ops->unmap_single(esp, ent->sense_dma,
					       SCSI_SENSE_BUFFERSIZE,
					       DMA_FROM_DEVICE);
			ent->sense_ptr = NULL;
		}

		/* Now that the state is unwound properly, put back onto
		 * the issue queue.  This command is no longer active.
		 */
		list_move(&ent->list, &esp->queued_cmds);
		esp->active_cmd = NULL;

		/* Return value ignored by caller, it directly invokes
		 * esp_reconnect().
		 */
		return 0;
	}

	if (esp->ireg == ESP_INTR_DC) {
		struct scsi_device *dev = cmd->device;

		/* Disconnect.  Make sure we re-negotiate sync and
		 * wide parameters if this target starts responding
		 * again in the future.
		 */
		esp->target[dev->id].flags |= ESP_TGT_CHECK_NEGO;

		scsi_esp_cmd(esp, ESP_CMD_ESEL);
		esp_cmd_is_done(esp, ent, cmd, (DID_BAD_TARGET << 16));
		return 1;
	}

	if (esp->ireg == (ESP_INTR_FDONE | ESP_INTR_BSERV)) {
		/* Selection successful.  On pre-FAST chips we have
		 * to do a NOP and possibly clean out the FIFO.
		 */
		if (esp->rev <= ESP236) {
			int fcnt = esp_read8(ESP_FFLAGS) & ESP_FF_FBYTES;

			scsi_esp_cmd(esp, ESP_CMD_NULL);

			if (!fcnt &&
			    (!esp->prev_soff ||
			     ((esp->sreg & ESP_STAT_PMASK) != ESP_DIP)))
				esp_flush_fifo(esp);
		}

		/* If we are doing a slow command, negotiation, etc.
		 * we'll do the right thing as we transition to the
		 * next phase.
		 */
		esp_event(esp, ESP_EVENT_CHECK_PHASE);
		return 0;
	}

	printk("ESP: Unexpected selection completion ireg[%x].\n",
	       esp->ireg);
	esp_schedule_reset(esp);
	return 0;
}

static int esp_data_bytes_sent(struct esp *esp, struct esp_cmd_entry *ent,
			       struct scsi_cmnd *cmd)
{
	int fifo_cnt, ecount, bytes_sent, flush_fifo;

	fifo_cnt = esp_read8(ESP_FFLAGS) & ESP_FF_FBYTES;
	if (esp->prev_cfg3 & ESP_CONFIG3_EWIDE)
		fifo_cnt <<= 1;

	ecount = 0;
	if (!(esp->sreg & ESP_STAT_TCNT)) {
		ecount = ((unsigned int)esp_read8(ESP_TCLOW) |
			  (((unsigned int)esp_read8(ESP_TCMED)) << 8));
		if (esp->rev == FASHME)
			ecount |= ((unsigned int)esp_read8(FAS_RLO)) << 16;
	}

	bytes_sent = esp->data_dma_len;
	bytes_sent -= ecount;

	if (!(ent->flags & ESP_CMD_FLAG_WRITE))
		bytes_sent -= fifo_cnt;

	flush_fifo = 0;
	if (!esp->prev_soff) {
		/* Synchronous data transfer, always flush fifo. */
		flush_fifo = 1;
	} else {
		if (esp->rev == ESP100) {
			u32 fflags, phase;

			/* ESP100 has a chip bug where in the synchronous data
			 * phase it can mistake a final long REQ pulse from the
			 * target as an extra data byte.  Fun.
			 *
			 * To detect this case we resample the status register
			 * and fifo flags.  If we're still in a data phase and
			 * we see spurious chunks in the fifo, we return error
			 * to the caller which should reset and set things up
			 * such that we only try future transfers to this
			 * target in synchronous mode.
			 */
			esp->sreg = esp_read8(ESP_STATUS);
			phase = esp->sreg & ESP_STAT_PMASK;
			fflags = esp_read8(ESP_FFLAGS);

			if ((phase == ESP_DOP &&
			     (fflags & ESP_FF_ONOTZERO)) ||
			    (phase == ESP_DIP &&
			     (fflags & ESP_FF_FBYTES)))
				return -1;
		}
		if (!(ent->flags & ESP_CMD_FLAG_WRITE))
			flush_fifo = 1;
	}

	if (flush_fifo)
		esp_flush_fifo(esp);

	return bytes_sent;
}

static void esp_setsync(struct esp *esp, struct esp_target_data *tp,
			u8 scsi_period, u8 scsi_offset,
			u8 esp_stp, u8 esp_soff)
{
	spi_period(tp->starget) = scsi_period;
	spi_offset(tp->starget) = scsi_offset;
	spi_width(tp->starget) = (tp->flags & ESP_TGT_WIDE) ? 1 : 0;

	if (esp_soff) {
		esp_stp &= 0x1f;
		esp_soff |= esp->radelay;
		if (esp->rev >= FAS236) {
			u8 bit = ESP_CONFIG3_FSCSI;
			if (esp->rev >= FAS100A)
				bit = ESP_CONFIG3_FAST;

			if (scsi_period < 50) {
				if (esp->rev == FASHME)
					esp_soff &= ~esp->radelay;
				tp->esp_config3 |= bit;
			} else {
				tp->esp_config3 &= ~bit;
			}
			esp->prev_cfg3 = tp->esp_config3;
			esp_write8(esp->prev_cfg3, ESP_CFG3);
		}
	}

	tp->esp_period = esp->prev_stp = esp_stp;
	tp->esp_offset = esp->prev_soff = esp_soff;

	esp_write8(esp_soff, ESP_SOFF);
	esp_write8(esp_stp, ESP_STP);

	tp->flags &= ~(ESP_TGT_NEGO_SYNC | ESP_TGT_CHECK_NEGO);

	spi_display_xfer_agreement(tp->starget);
}

static void esp_msgin_reject(struct esp *esp)
{
	struct esp_cmd_entry *ent = esp->active_cmd;
	struct scsi_cmnd *cmd = ent->cmd;
	struct esp_target_data *tp;
	int tgt;

	tgt = cmd->device->id;
	tp = &esp->target[tgt];

	if (tp->flags & ESP_TGT_NEGO_WIDE) {
		tp->flags &= ~(ESP_TGT_NEGO_WIDE | ESP_TGT_WIDE);

		if (!esp_need_to_nego_sync(tp)) {
			tp->flags &= ~ESP_TGT_CHECK_NEGO;
			scsi_esp_cmd(esp, ESP_CMD_RATN);
		} else {
			esp->msg_out_len =
				spi_populate_sync_msg(&esp->msg_out[0],
						      tp->nego_goal_period,
						      tp->nego_goal_offset);
			tp->flags |= ESP_TGT_NEGO_SYNC;
			scsi_esp_cmd(esp, ESP_CMD_SATN);
		}
		return;
	}

	if (tp->flags & ESP_TGT_NEGO_SYNC) {
		tp->flags &= ~(ESP_TGT_NEGO_SYNC | ESP_TGT_CHECK_NEGO);
		tp->esp_period = 0;
		tp->esp_offset = 0;
		esp_setsync(esp, tp, 0, 0, 0, 0);
		scsi_esp_cmd(esp, ESP_CMD_RATN);
		return;
	}

	esp->msg_out[0] = ABORT_TASK_SET;
	esp->msg_out_len = 1;
	scsi_esp_cmd(esp, ESP_CMD_SATN);
}

static void esp_msgin_sdtr(struct esp *esp, struct esp_target_data *tp)
{
	u8 period = esp->msg_in[3];
	u8 offset = esp->msg_in[4];
	u8 stp;

	if (!(tp->flags & ESP_TGT_NEGO_SYNC))
		goto do_reject;

	if (offset > 15)
		goto do_reject;

	if (offset) {
		int one_clock;

		if (period > esp->max_period) {
			period = offset = 0;
			goto do_sdtr;
		}
		if (period < esp->min_period)
			goto do_reject;

		one_clock = esp->ccycle / 1000;
		stp = DIV_ROUND_UP(period << 2, one_clock);
		if (stp && esp->rev >= FAS236) {
			if (stp >= 50)
				stp--;
		}
	} else {
		stp = 0;
	}

	esp_setsync(esp, tp, period, offset, stp, offset);
	return;

do_reject:
	esp->msg_out[0] = MESSAGE_REJECT;
	esp->msg_out_len = 1;
	scsi_esp_cmd(esp, ESP_CMD_SATN);
	return;

do_sdtr:
	tp->nego_goal_period = period;
	tp->nego_goal_offset = offset;
	esp->msg_out_len =
		spi_populate_sync_msg(&esp->msg_out[0],
				      tp->nego_goal_period,
				      tp->nego_goal_offset);
	scsi_esp_cmd(esp, ESP_CMD_SATN);
}

static void esp_msgin_wdtr(struct esp *esp, struct esp_target_data *tp)
{
	int size = 8 << esp->msg_in[3];
	u8 cfg3;

	if (esp->rev != FASHME)
		goto do_reject;

	if (size != 8 && size != 16)
		goto do_reject;

	if (!(tp->flags & ESP_TGT_NEGO_WIDE))
		goto do_reject;

	cfg3 = tp->esp_config3;
	if (size == 16) {
		tp->flags |= ESP_TGT_WIDE;
		cfg3 |= ESP_CONFIG3_EWIDE;
	} else {
		tp->flags &= ~ESP_TGT_WIDE;
		cfg3 &= ~ESP_CONFIG3_EWIDE;
	}
	tp->esp_config3 = cfg3;
	esp->prev_cfg3 = cfg3;
	esp_write8(cfg3, ESP_CFG3);

	tp->flags &= ~ESP_TGT_NEGO_WIDE;

	spi_period(tp->starget) = 0;
	spi_offset(tp->starget) = 0;
	if (!esp_need_to_nego_sync(tp)) {
		tp->flags &= ~ESP_TGT_CHECK_NEGO;
		scsi_esp_cmd(esp, ESP_CMD_RATN);
	} else {
		esp->msg_out_len =
			spi_populate_sync_msg(&esp->msg_out[0],
					      tp->nego_goal_period,
					      tp->nego_goal_offset);
		tp->flags |= ESP_TGT_NEGO_SYNC;
		scsi_esp_cmd(esp, ESP_CMD_SATN);
	}
	return;

do_reject:
	esp->msg_out[0] = MESSAGE_REJECT;
	esp->msg_out_len = 1;
	scsi_esp_cmd(esp, ESP_CMD_SATN);
}

static void esp_msgin_extended(struct esp *esp)
{
	struct esp_cmd_entry *ent = esp->active_cmd;
	struct scsi_cmnd *cmd = ent->cmd;
	struct esp_target_data *tp;
	int tgt = cmd->device->id;

	tp = &esp->target[tgt];
	if (esp->msg_in[2] == EXTENDED_SDTR) {
		esp_msgin_sdtr(esp, tp);
		return;
	}
	if (esp->msg_in[2] == EXTENDED_WDTR) {
		esp_msgin_wdtr(esp, tp);
		return;
	}

	printk("ESP: Unexpected extended msg type %x\n",
	       esp->msg_in[2]);

	esp->msg_out[0] = ABORT_TASK_SET;
	esp->msg_out_len = 1;
	scsi_esp_cmd(esp, ESP_CMD_SATN);
}

/* Analyze msgin bytes received from target so far.  Return non-zero
 * if there are more bytes needed to complete the message.
 */
static int esp_msgin_process(struct esp *esp)
{
	u8 msg0 = esp->msg_in[0];
	int len = esp->msg_in_len;

	if (msg0 & 0x80) {
		/* Identify */
		printk("ESP: Unexpected msgin identify\n");
		return 0;
	}

	switch (msg0) {
	case EXTENDED_MESSAGE:
		if (len == 1)
			return 1;
		if (len < esp->msg_in[1] + 2)
			return 1;
		esp_msgin_extended(esp);
		return 0;

	case IGNORE_WIDE_RESIDUE: {
		struct esp_cmd_entry *ent;
		struct esp_cmd_priv *spriv;
		if (len == 1)
			return 1;

		if (esp->msg_in[1] != 1)
			goto do_reject;

		ent = esp->active_cmd;
		spriv = ESP_CMD_PRIV(ent->cmd);

		if (spriv->cur_residue == sg_dma_len(spriv->cur_sg)) {
			spriv->cur_sg--;
			spriv->cur_residue = 1;
		} else
			spriv->cur_residue++;
		spriv->tot_residue++;
		return 0;
	}
	case NOP:
		return 0;
	case RESTORE_POINTERS:
		esp_restore_pointers(esp, esp->active_cmd);
		return 0;
	case SAVE_POINTERS:
		esp_save_pointers(esp, esp->active_cmd);
		return 0;

	case COMMAND_COMPLETE:
	case DISCONNECT: {
		struct esp_cmd_entry *ent = esp->active_cmd;

		ent->message = msg0;
		esp_event(esp, ESP_EVENT_FREE_BUS);
		esp->flags |= ESP_FLAG_QUICKIRQ_CHECK;
		return 0;
	}
	case MESSAGE_REJECT:
		esp_msgin_reject(esp);
		return 0;

	default:
	do_reject:
		esp->msg_out[0] = MESSAGE_REJECT;
		esp->msg_out_len = 1;
		scsi_esp_cmd(esp, ESP_CMD_SATN);
		return 0;
	}
}

static int esp_process_event(struct esp *esp)
{
	int write;

again:
	write = 0;
	switch (esp->event) {
	case ESP_EVENT_CHECK_PHASE:
		switch (esp->sreg & ESP_STAT_PMASK) {
		case ESP_DOP:
			esp_event(esp, ESP_EVENT_DATA_OUT);
			break;
		case ESP_DIP:
			esp_event(esp, ESP_EVENT_DATA_IN);
			break;
		case ESP_STATP:
			esp_flush_fifo(esp);
			scsi_esp_cmd(esp, ESP_CMD_ICCSEQ);
			esp_event(esp, ESP_EVENT_STATUS);
			esp->flags |= ESP_FLAG_QUICKIRQ_CHECK;
			return 1;

		case ESP_MOP:
			esp_event(esp, ESP_EVENT_MSGOUT);
			break;

		case ESP_MIP:
			esp_event(esp, ESP_EVENT_MSGIN);
			break;

		case ESP_CMDP:
			esp_event(esp, ESP_EVENT_CMD_START);
			break;

		default:
			printk("ESP: Unexpected phase, sreg=%02x\n",
			       esp->sreg);
			esp_schedule_reset(esp);
			return 0;
		}
		goto again;
		break;

	case ESP_EVENT_DATA_IN:
		write = 1;
		/* fallthru */

	case ESP_EVENT_DATA_OUT: {
		struct esp_cmd_entry *ent = esp->active_cmd;
		struct scsi_cmnd *cmd = ent->cmd;
		dma_addr_t dma_addr = esp_cur_dma_addr(ent, cmd);
		unsigned int dma_len = esp_cur_dma_len(ent, cmd);

		if (esp->rev == ESP100)
			scsi_esp_cmd(esp, ESP_CMD_NULL);

		if (write)
			ent->flags |= ESP_CMD_FLAG_WRITE;
		else
			ent->flags &= ~ESP_CMD_FLAG_WRITE;

		if (esp->ops->dma_length_limit)
			dma_len = esp->ops->dma_length_limit(esp, dma_addr,
							     dma_len);
		else
			dma_len = esp_dma_length_limit(esp, dma_addr, dma_len);

		esp->data_dma_len = dma_len;

		if (!dma_len) {
			printk(KERN_ERR PFX "esp%d: DMA length is zero!\n",
			       esp->host->unique_id);
			printk(KERN_ERR PFX "esp%d: cur adr[%08llx] len[%08x]\n",
			       esp->host->unique_id,
			       (unsigned long long)esp_cur_dma_addr(ent, cmd),
			       esp_cur_dma_len(ent, cmd));
			esp_schedule_reset(esp);
			return 0;
		}

		esp_log_datastart("ESP: start data addr[%08llx] len[%u] "
				  "write(%d)\n",
				  (unsigned long long)dma_addr, dma_len, write);

		esp->ops->send_dma_cmd(esp, dma_addr, dma_len, dma_len,
				       write, ESP_CMD_DMA | ESP_CMD_TI);
		esp_event(esp, ESP_EVENT_DATA_DONE);
		break;
	}
	case ESP_EVENT_DATA_DONE: {
		struct esp_cmd_entry *ent = esp->active_cmd;
		struct scsi_cmnd *cmd = ent->cmd;
		int bytes_sent;

		if (esp->ops->dma_error(esp)) {
			printk("ESP: data done, DMA error, resetting\n");
			esp_schedule_reset(esp);
			return 0;
		}

		if (ent->flags & ESP_CMD_FLAG_WRITE) {
			/* XXX parity errors, etc. XXX */

			esp->ops->dma_drain(esp);
		}
		esp->ops->dma_invalidate(esp);

		if (esp->ireg != ESP_INTR_BSERV) {
			/* We should always see exactly a bus-service
			 * interrupt at the end of a successful transfer.
			 */
			printk("ESP: data done, not BSERV, resetting\n");
			esp_schedule_reset(esp);
			return 0;
		}

		bytes_sent = esp_data_bytes_sent(esp, ent, cmd);

		esp_log_datadone("ESP: data done flgs[%x] sent[%d]\n",
				 ent->flags, bytes_sent);

		if (bytes_sent < 0) {
			/* XXX force sync mode for this target XXX */
			esp_schedule_reset(esp);
			return 0;
		}

		esp_advance_dma(esp, ent, cmd, bytes_sent);
		esp_event(esp, ESP_EVENT_CHECK_PHASE);
		goto again;
	}

	case ESP_EVENT_STATUS: {
		struct esp_cmd_entry *ent = esp->active_cmd;

		if (esp->ireg & ESP_INTR_FDONE) {
			ent->status = esp_read8(ESP_FDATA);
			ent->message = esp_read8(ESP_FDATA);
			scsi_esp_cmd(esp, ESP_CMD_MOK);
		} else if (esp->ireg == ESP_INTR_BSERV) {
			ent->status = esp_read8(ESP_FDATA);
			ent->message = 0xff;
			esp_event(esp, ESP_EVENT_MSGIN);
			return 0;
		}

		if (ent->message != COMMAND_COMPLETE) {
			printk("ESP: Unexpected message %x in status\n",
			       ent->message);
			esp_schedule_reset(esp);
			return 0;
		}

		esp_event(esp, ESP_EVENT_FREE_BUS);
		esp->flags |= ESP_FLAG_QUICKIRQ_CHECK;
		break;
	}
	case ESP_EVENT_FREE_BUS: {
		struct esp_cmd_entry *ent = esp->active_cmd;
		struct scsi_cmnd *cmd = ent->cmd;

		if (ent->message == COMMAND_COMPLETE ||
		    ent->message == DISCONNECT)
			scsi_esp_cmd(esp, ESP_CMD_ESEL);

		if (ent->message == COMMAND_COMPLETE) {
			esp_log_cmddone("ESP: Command done status[%x] "
					"message[%x]\n",
					ent->status, ent->message);
			if (ent->status == SAM_STAT_TASK_SET_FULL)
				esp_event_queue_full(esp, ent);

			if (ent->status == SAM_STAT_CHECK_CONDITION &&
			    !(ent->flags & ESP_CMD_FLAG_AUTOSENSE)) {
				ent->flags |= ESP_CMD_FLAG_AUTOSENSE;
				esp_autosense(esp, ent);
			} else {
				esp_cmd_is_done(esp, ent, cmd,
						compose_result(ent->status,
							       ent->message,
							       DID_OK));
			}
		} else if (ent->message == DISCONNECT) {
			esp_log_disconnect("ESP: Disconnecting tgt[%d] "
					   "tag[%x:%x]\n",
					   cmd->device->id,
					   ent->tag[0], ent->tag[1]);

			esp->active_cmd = NULL;
			esp_maybe_execute_command(esp);
		} else {
			printk("ESP: Unexpected message %x in freebus\n",
			       ent->message);
			esp_schedule_reset(esp);
			return 0;
		}
		if (esp->active_cmd)
			esp->flags |= ESP_FLAG_QUICKIRQ_CHECK;
		break;
	}
	case ESP_EVENT_MSGOUT: {
		scsi_esp_cmd(esp, ESP_CMD_FLUSH);

		if (esp_debug & ESP_DEBUG_MSGOUT) {
			int i;
			printk("ESP: Sending message [ ");
			for (i = 0; i < esp->msg_out_len; i++)
				printk("%02x ", esp->msg_out[i]);
			printk("]\n");
		}

		if (esp->rev == FASHME) {
			int i;

			/* Always use the fifo.  */
			for (i = 0; i < esp->msg_out_len; i++) {
				esp_write8(esp->msg_out[i], ESP_FDATA);
				esp_write8(0, ESP_FDATA);
			}
			scsi_esp_cmd(esp, ESP_CMD_TI);
		} else {
			if (esp->msg_out_len == 1) {
				esp_write8(esp->msg_out[0], ESP_FDATA);
				scsi_esp_cmd(esp, ESP_CMD_TI);
			} else {
				/* Use DMA. */
				memcpy(esp->command_block,
				       esp->msg_out,
				       esp->msg_out_len);

				esp->ops->send_dma_cmd(esp,
						       esp->command_block_dma,
						       esp->msg_out_len,
						       esp->msg_out_len,
						       0,
						       ESP_CMD_DMA|ESP_CMD_TI);
			}
		}
		esp_event(esp, ESP_EVENT_MSGOUT_DONE);
		break;
	}
	case ESP_EVENT_MSGOUT_DONE:
		if (esp->rev == FASHME) {
			scsi_esp_cmd(esp, ESP_CMD_FLUSH);
		} else {
			if (esp->msg_out_len > 1)
				esp->ops->dma_invalidate(esp);
		}

		if (!(esp->ireg & ESP_INTR_DC)) {
			if (esp->rev != FASHME)
				scsi_esp_cmd(esp, ESP_CMD_NULL);
		}
		esp_event(esp, ESP_EVENT_CHECK_PHASE);
		goto again;
	case ESP_EVENT_MSGIN:
		if (esp->ireg & ESP_INTR_BSERV) {
			if (esp->rev == FASHME) {
				if (!(esp_read8(ESP_STATUS2) &
				      ESP_STAT2_FEMPTY))
					scsi_esp_cmd(esp, ESP_CMD_FLUSH);
			} else {
				scsi_esp_cmd(esp, ESP_CMD_FLUSH);
				if (esp->rev == ESP100)
					scsi_esp_cmd(esp, ESP_CMD_NULL);
			}
			scsi_esp_cmd(esp, ESP_CMD_TI);
			esp->flags |= ESP_FLAG_QUICKIRQ_CHECK;
			return 1;
		}
		if (esp->ireg & ESP_INTR_FDONE) {
			u8 val;

			if (esp->rev == FASHME)
				val = esp->fifo[0];
			else
				val = esp_read8(ESP_FDATA);
			esp->msg_in[esp->msg_in_len++] = val;

			esp_log_msgin("ESP: Got msgin byte %x\n", val);

			if (!esp_msgin_process(esp))
				esp->msg_in_len = 0;

			if (esp->rev == FASHME)
				scsi_esp_cmd(esp, ESP_CMD_FLUSH);

			scsi_esp_cmd(esp, ESP_CMD_MOK);

			if (esp->event != ESP_EVENT_FREE_BUS)
				esp_event(esp, ESP_EVENT_CHECK_PHASE);
		} else {
			printk("ESP: MSGIN neither BSERV not FDON, resetting");
			esp_schedule_reset(esp);
			return 0;
		}
		break;
	case ESP_EVENT_CMD_START:
		memcpy(esp->command_block, esp->cmd_bytes_ptr,
		       esp->cmd_bytes_left);
		if (esp->rev == FASHME)
			scsi_esp_cmd(esp, ESP_CMD_FLUSH);
		esp->ops->send_dma_cmd(esp, esp->command_block_dma,
				       esp->cmd_bytes_left, 16, 0,
				       ESP_CMD_DMA | ESP_CMD_TI);
		esp_event(esp, ESP_EVENT_CMD_DONE);
		esp->flags |= ESP_FLAG_QUICKIRQ_CHECK;
		break;
	case ESP_EVENT_CMD_DONE:
		esp->ops->dma_invalidate(esp);
		if (esp->ireg & ESP_INTR_BSERV) {
			esp_event(esp, ESP_EVENT_CHECK_PHASE);
			goto again;
		}
		esp_schedule_reset(esp);
		return 0;
		break;

	case ESP_EVENT_RESET:
		scsi_esp_cmd(esp, ESP_CMD_RS);
		break;

	default:
		printk("ESP: Unexpected event %x, resetting\n",
		       esp->event);
		esp_schedule_reset(esp);
		return 0;
		break;
	}
	return 1;
}

static void esp_reset_cleanup_one(struct esp *esp, struct esp_cmd_entry *ent)
{
	struct scsi_cmnd *cmd = ent->cmd;

	esp_unmap_dma(esp, cmd);
	esp_free_lun_tag(ent, cmd->device->hostdata);
	cmd->result = DID_RESET << 16;

	if (ent->flags & ESP_CMD_FLAG_AUTOSENSE) {
		esp->ops->unmap_single(esp, ent->sense_dma,
				       SCSI_SENSE_BUFFERSIZE, DMA_FROM_DEVICE);
		ent->sense_ptr = NULL;
	}

	cmd->scsi_done(cmd);
	list_del(&ent->list);
	esp_put_ent(esp, ent);
}

static void esp_clear_hold(struct scsi_device *dev, void *data)
{
	struct esp_lun_data *lp = dev->hostdata;

	BUG_ON(lp->num_tagged);
	lp->hold = 0;
}

static void esp_reset_cleanup(struct esp *esp)
{
	struct esp_cmd_entry *ent, *tmp;
	int i;

	list_for_each_entry_safe(ent, tmp, &esp->queued_cmds, list) {
		struct scsi_cmnd *cmd = ent->cmd;

		list_del(&ent->list);
		cmd->result = DID_RESET << 16;
		cmd->scsi_done(cmd);
		esp_put_ent(esp, ent);
	}

	list_for_each_entry_safe(ent, tmp, &esp->active_cmds, list) {
		if (ent == esp->active_cmd)
			esp->active_cmd = NULL;
		esp_reset_cleanup_one(esp, ent);
	}

	BUG_ON(esp->active_cmd != NULL);

	/* Force renegotiation of sync/wide transfers.  */
	for (i = 0; i < ESP_MAX_TARGET; i++) {
		struct esp_target_data *tp = &esp->target[i];

		tp->esp_period = 0;
		tp->esp_offset = 0;
		tp->esp_config3 &= ~(ESP_CONFIG3_EWIDE |
				     ESP_CONFIG3_FSCSI |
				     ESP_CONFIG3_FAST);
		tp->flags &= ~ESP_TGT_WIDE;
		tp->flags |= ESP_TGT_CHECK_NEGO;

		if (tp->starget)
			__starget_for_each_device(tp->starget, NULL,
						  esp_clear_hold);
	}
	esp->flags &= ~ESP_FLAG_RESETTING;
}

/* Runs under host->lock */
static void __esp_interrupt(struct esp *esp)
{
	int finish_reset, intr_done;
	u8 phase;

	esp->sreg = esp_read8(ESP_STATUS);

	if (esp->flags & ESP_FLAG_RESETTING) {
		finish_reset = 1;
	} else {
		if (esp_check_gross_error(esp))
			return;

		finish_reset = esp_check_spur_intr(esp);
		if (finish_reset < 0)
			return;
	}

	esp->ireg = esp_read8(ESP_INTRPT);

	if (esp->ireg & ESP_INTR_SR)
		finish_reset = 1;

	if (finish_reset) {
		esp_reset_cleanup(esp);
		if (esp->eh_reset) {
			complete(esp->eh_reset);
			esp->eh_reset = NULL;
		}
		return;
	}

	phase = (esp->sreg & ESP_STAT_PMASK);
	if (esp->rev == FASHME) {
		if (((phase != ESP_DIP && phase != ESP_DOP) &&
		     esp->select_state == ESP_SELECT_NONE &&
		     esp->event != ESP_EVENT_STATUS &&
		     esp->event != ESP_EVENT_DATA_DONE) ||
		    (esp->ireg & ESP_INTR_RSEL)) {
			esp->sreg2 = esp_read8(ESP_STATUS2);
			if (!(esp->sreg2 & ESP_STAT2_FEMPTY) ||
			    (esp->sreg2 & ESP_STAT2_F1BYTE))
				hme_read_fifo(esp);
		}
	}

	esp_log_intr("ESP: intr sreg[%02x] seqreg[%02x] "
		     "sreg2[%02x] ireg[%02x]\n",
		     esp->sreg, esp->seqreg, esp->sreg2, esp->ireg);

	intr_done = 0;

	if (esp->ireg & (ESP_INTR_S | ESP_INTR_SATN | ESP_INTR_IC)) {
		printk("ESP: unexpected IREG %02x\n", esp->ireg);
		if (esp->ireg & ESP_INTR_IC)
			esp_dump_cmd_log(esp);

		esp_schedule_reset(esp);
	} else {
		if (!(esp->ireg & ESP_INTR_RSEL)) {
			/* Some combination of FDONE, BSERV, DC.  */
			if (esp->select_state != ESP_SELECT_NONE)
				intr_done = esp_finish_select(esp);
		} else if (esp->ireg & ESP_INTR_RSEL) {
			if (esp->active_cmd)
				(void) esp_finish_select(esp);
			intr_done = esp_reconnect(esp);
		}
	}
	while (!intr_done)
		intr_done = esp_process_event(esp);
}

irqreturn_t scsi_esp_intr(int irq, void *dev_id)
{
	struct esp *esp = dev_id;
	unsigned long flags;
	irqreturn_t ret;

	spin_lock_irqsave(esp->host->host_lock, flags);
	ret = IRQ_NONE;
	if (esp->ops->irq_pending(esp)) {
		ret = IRQ_HANDLED;
		for (;;) {
			int i;

			__esp_interrupt(esp);
			if (!(esp->flags & ESP_FLAG_QUICKIRQ_CHECK))
				break;
			esp->flags &= ~ESP_FLAG_QUICKIRQ_CHECK;

			for (i = 0; i < ESP_QUICKIRQ_LIMIT; i++) {
				if (esp->ops->irq_pending(esp))
					break;
			}
			if (i == ESP_QUICKIRQ_LIMIT)
				break;
		}
	}
	spin_unlock_irqrestore(esp->host->host_lock, flags);

	return ret;
}
EXPORT_SYMBOL(scsi_esp_intr);

static void esp_get_revision(struct esp *esp)
{
	u8 val;

	esp->config1 = (ESP_CONFIG1_PENABLE | (esp->scsi_id & 7));
	esp->config2 = (ESP_CONFIG2_SCSI2ENAB | ESP_CONFIG2_REGPARITY);
	esp_write8(esp->config2, ESP_CFG2);

	val = esp_read8(ESP_CFG2);
	val &= ~ESP_CONFIG2_MAGIC;
	if (val != (ESP_CONFIG2_SCSI2ENAB | ESP_CONFIG2_REGPARITY)) {
		/* If what we write to cfg2 does not come back, cfg2 is not
		 * implemented, therefore this must be a plain esp100.
		 */
		esp->rev = ESP100;
	} else {
		esp->config2 = 0;
		esp_set_all_config3(esp, 5);
		esp->prev_cfg3 = 5;
		esp_write8(esp->config2, ESP_CFG2);
		esp_write8(0, ESP_CFG3);
		esp_write8(esp->prev_cfg3, ESP_CFG3);

		val = esp_read8(ESP_CFG3);
		if (val != 5) {
			/* The cfg2 register is implemented, however
			 * cfg3 is not, must be esp100a.
			 */
			esp->rev = ESP100A;
		} else {
			esp_set_all_config3(esp, 0);
			esp->prev_cfg3 = 0;
			esp_write8(esp->prev_cfg3, ESP_CFG3);

			/* All of cfg{1,2,3} implemented, must be one of
			 * the fas variants, figure out which one.
			 */
			if (esp->cfact == 0 || esp->cfact > ESP_CCF_F5) {
				esp->rev = FAST;
				esp->sync_defp = SYNC_DEFP_FAST;
			} else {
				esp->rev = ESP236;
			}
			esp->config2 = 0;
			esp_write8(esp->config2, ESP_CFG2);
		}
	}
}

static void esp_init_swstate(struct esp *esp)
{
	int i;

	INIT_LIST_HEAD(&esp->queued_cmds);
	INIT_LIST_HEAD(&esp->active_cmds);
	INIT_LIST_HEAD(&esp->esp_cmd_pool);

	/* Start with a clear state, domain validation (via ->slave_configure,
	 * spi_dv_device()) will attempt to enable SYNC, WIDE, and tagged
	 * commands.
	 */
	for (i = 0 ; i < ESP_MAX_TARGET; i++) {
		esp->target[i].flags = 0;
		esp->target[i].nego_goal_period = 0;
		esp->target[i].nego_goal_offset = 0;
		esp->target[i].nego_goal_width = 0;
		esp->target[i].nego_goal_tags = 0;
	}
}

/* This places the ESP into a known state at boot time. */
static void esp_bootup_reset(struct esp *esp)
{
	u8 val;

	/* Reset the DMA */
	esp->ops->reset_dma(esp);

	/* Reset the ESP */
	esp_reset_esp(esp);

	/* Reset the SCSI bus, but tell ESP not to generate an irq */
	val = esp_read8(ESP_CFG1);
	val |= ESP_CONFIG1_SRRDISAB;
	esp_write8(val, ESP_CFG1);

	scsi_esp_cmd(esp, ESP_CMD_RS);
	udelay(400);

	esp_write8(esp->config1, ESP_CFG1);

	/* Eat any bitrot in the chip and we are done... */
	esp_read8(ESP_INTRPT);
}

static void esp_set_clock_params(struct esp *esp)
{
	int fhz;
	u8 ccf;

	/* This is getting messy but it has to be done correctly or else
	 * you get weird behavior all over the place.  We are trying to
	 * basically figure out three pieces of information.
	 *
	 * a) Clock Conversion Factor
	 *
	 *    This is a representation of the input crystal clock frequency
	 *    going into the ESP on this machine.  Any operation whose timing
	 *    is longer than 400ns depends on this value being correct.  For
	 *    example, you'll get blips for arbitration/selection during high
	 *    load or with multiple targets if this is not set correctly.
	 *
	 * b) Selection Time-Out
	 *
	 *    The ESP isn't very bright and will arbitrate for the bus and try
	 *    to select a target forever if you let it.  This value tells the
	 *    ESP when it has taken too long to negotiate and that it should
	 *    interrupt the CPU so we can see what happened.  The value is
	 *    computed as follows (from NCR/Symbios chip docs).
	 *
	 *          (Time Out Period) *  (Input Clock)
	 *    STO = ----------------------------------
	 *          (8192) * (Clock Conversion Factor)
	 *
	 *    We use a time out period of 250ms (ESP_BUS_TIMEOUT).
	 *
	 * c) Imperical constants for synchronous offset and transfer period
         *    register values
	 *
	 *    This entails the smallest and largest sync period we could ever
	 *    handle on this ESP.
	 */
	fhz = esp->cfreq;

	ccf = ((fhz / 1000000) + 4) / 5;
	if (ccf == 1)
		ccf = 2;

	/* If we can't find anything reasonable, just assume 20MHZ.
	 * This is the clock frequency of the older sun4c's where I've
	 * been unable to find the clock-frequency PROM property.  All
	 * other machines provide useful values it seems.
	 */
	if (fhz <= 5000000 || ccf < 1 || ccf > 8) {
		fhz = 20000000;
		ccf = 4;
	}

	esp->cfact = (ccf == 8 ? 0 : ccf);
	esp->cfreq = fhz;
	esp->ccycle = ESP_HZ_TO_CYCLE(fhz);
	esp->ctick = ESP_TICK(ccf, esp->ccycle);
	esp->neg_defp = ESP_NEG_DEFP(fhz, ccf);
	esp->sync_defp = SYNC_DEFP_SLOW;
}

static const char *esp_chip_names[] = {
	"ESP100",
	"ESP100A",
	"ESP236",
	"FAS236",
	"FAS100A",
	"FAST",
	"FASHME",
};

static struct scsi_transport_template *esp_transport_template;

int scsi_esp_register(struct esp *esp, struct device *dev)
{
	static int instance;
	int err;

	esp->host->transportt = esp_transport_template;
	esp->host->max_lun = ESP_MAX_LUN;
	esp->host->cmd_per_lun = 2;
	esp->host->unique_id = instance;

	esp_set_clock_params(esp);

	esp_get_revision(esp);

	esp_init_swstate(esp);

	esp_bootup_reset(esp);

	printk(KERN_INFO PFX "esp%u, regs[%1p:%1p] irq[%u]\n",
	       esp->host->unique_id, esp->regs, esp->dma_regs,
	       esp->host->irq);
	printk(KERN_INFO PFX "esp%u is a %s, %u MHz (ccf=%u), SCSI ID %u\n",
	       esp->host->unique_id, esp_chip_names[esp->rev],
	       esp->cfreq / 1000000, esp->cfact, esp->scsi_id);

	/* Let the SCSI bus reset settle. */
	ssleep(esp_bus_reset_settle);

	err = scsi_add_host(esp->host, dev);
	if (err)
		return err;

	instance++;

	scsi_scan_host(esp->host);

	return 0;
}
EXPORT_SYMBOL(scsi_esp_register);

void scsi_esp_unregister(struct esp *esp)
{
	scsi_remove_host(esp->host);
}
EXPORT_SYMBOL(scsi_esp_unregister);

static int esp_target_alloc(struct scsi_target *starget)
{
	struct esp *esp = shost_priv(dev_to_shost(&starget->dev));
	struct esp_target_data *tp = &esp->target[starget->id];

	tp->starget = starget;

	return 0;
}

static void esp_target_destroy(struct scsi_target *starget)
{
	struct esp *esp = shost_priv(dev_to_shost(&starget->dev));
	struct esp_target_data *tp = &esp->target[starget->id];

	tp->starget = NULL;
}

static int esp_slave_alloc(struct scsi_device *dev)
{
	struct esp *esp = shost_priv(dev->host);
	struct esp_target_data *tp = &esp->target[dev->id];
	struct esp_lun_data *lp;

	lp = kzalloc(sizeof(*lp), GFP_KERNEL);
	if (!lp)
		return -ENOMEM;
	dev->hostdata = lp;

	spi_min_period(tp->starget) = esp->min_period;
	spi_max_offset(tp->starget) = 15;

	if (esp->flags & ESP_FLAG_WIDE_CAPABLE)
		spi_max_width(tp->starget) = 1;
	else
		spi_max_width(tp->starget) = 0;

	return 0;
}

static int esp_slave_configure(struct scsi_device *dev)
{
	struct esp *esp = shost_priv(dev->host);
	struct esp_target_data *tp = &esp->target[dev->id];
	int goal_tags, queue_depth;

	goal_tags = 0;

	if (dev->tagged_supported) {
		/* XXX make this configurable somehow XXX */
		goal_tags = ESP_DEFAULT_TAGS;

		if (goal_tags > ESP_MAX_TAG)
			goal_tags = ESP_MAX_TAG;
	}

	queue_depth = goal_tags;
	if (queue_depth < dev->host->cmd_per_lun)
		queue_depth = dev->host->cmd_per_lun;

	if (goal_tags) {
		scsi_set_tag_type(dev, MSG_ORDERED_TAG);
		scsi_activate_tcq(dev, queue_depth);
	} else {
		scsi_deactivate_tcq(dev, queue_depth);
	}
	tp->flags |= ESP_TGT_DISCONNECT;

	if (!spi_initial_dv(dev->sdev_target))
		spi_dv_device(dev);

	return 0;
}

static void esp_slave_destroy(struct scsi_device *dev)
{
	struct esp_lun_data *lp = dev->hostdata;

	kfree(lp);
	dev->hostdata = NULL;
}

static int esp_eh_abort_handler(struct scsi_cmnd *cmd)
{
	struct esp *esp = shost_priv(cmd->device->host);
	struct esp_cmd_entry *ent, *tmp;
	struct completion eh_done;
	unsigned long flags;

	/* XXX This helps a lot with debugging but might be a bit
	 * XXX much for the final driver.
	 */
	spin_lock_irqsave(esp->host->host_lock, flags);
	printk(KERN_ERR PFX "esp%d: Aborting command [%p:%02x]\n",
	       esp->host->unique_id, cmd, cmd->cmnd[0]);
	ent = esp->active_cmd;
	if (ent)
		printk(KERN_ERR PFX "esp%d: Current command [%p:%02x]\n",
		       esp->host->unique_id, ent->cmd, ent->cmd->cmnd[0]);
	list_for_each_entry(ent, &esp->queued_cmds, list) {
		printk(KERN_ERR PFX "esp%d: Queued command [%p:%02x]\n",
		       esp->host->unique_id, ent->cmd, ent->cmd->cmnd[0]);
	}
	list_for_each_entry(ent, &esp->active_cmds, list) {
		printk(KERN_ERR PFX "esp%d: Active command [%p:%02x]\n",
		       esp->host->unique_id, ent->cmd, ent->cmd->cmnd[0]);
	}
	esp_dump_cmd_log(esp);
	spin_unlock_irqrestore(esp->host->host_lock, flags);

	spin_lock_irqsave(esp->host->host_lock, flags);

	ent = NULL;
	list_for_each_entry(tmp, &esp->queued_cmds, list) {
		if (tmp->cmd == cmd) {
			ent = tmp;
			break;
		}
	}

	if (ent) {
		/* Easiest case, we didn't even issue the command
		 * yet so it is trivial to abort.
		 */
		list_del(&ent->list);

		cmd->result = DID_ABORT << 16;
		cmd->scsi_done(cmd);

		esp_put_ent(esp, ent);

		goto out_success;
	}

	init_completion(&eh_done);

	ent = esp->active_cmd;
	if (ent && ent->cmd == cmd) {
		/* Command is the currently active command on
		 * the bus.  If we already have an output message
		 * pending, no dice.
		 */
		if (esp->msg_out_len)
			goto out_failure;

		/* Send out an abort, encouraging the target to
		 * go to MSGOUT phase by asserting ATN.
		 */
		esp->msg_out[0] = ABORT_TASK_SET;
		esp->msg_out_len = 1;
		ent->eh_done = &eh_done;

		scsi_esp_cmd(esp, ESP_CMD_SATN);
	} else {
		/* The command is disconnected.  This is not easy to
		 * abort.  For now we fail and let the scsi error
		 * handling layer go try a scsi bus reset or host
		 * reset.
		 *
		 * What we could do is put together a scsi command
		 * solely for the purpose of sending an abort message
		 * to the target.  Coming up with all the code to
		 * cook up scsi commands, special case them everywhere,
		 * etc. is for questionable gain and it would be better
		 * if the generic scsi error handling layer could do at
		 * least some of that for us.
		 *
		 * Anyways this is an area for potential future improvement
		 * in this driver.
		 */
		goto out_failure;
	}

	spin_unlock_irqrestore(esp->host->host_lock, flags);

	if (!wait_for_completion_timeout(&eh_done, 5 * HZ)) {
		spin_lock_irqsave(esp->host->host_lock, flags);
		ent->eh_done = NULL;
		spin_unlock_irqrestore(esp->host->host_lock, flags);

		return FAILED;
	}

	return SUCCESS;

out_success:
	spin_unlock_irqrestore(esp->host->host_lock, flags);
	return SUCCESS;

out_failure:
	/* XXX This might be a good location to set ESP_TGT_BROKEN
	 * XXX since we know which target/lun in particular is
	 * XXX causing trouble.
	 */
	spin_unlock_irqrestore(esp->host->host_lock, flags);
	return FAILED;
}

static int esp_eh_bus_reset_handler(struct scsi_cmnd *cmd)
{
	struct esp *esp = shost_priv(cmd->device->host);
	struct completion eh_reset;
	unsigned long flags;

	init_completion(&eh_reset);

	spin_lock_irqsave(esp->host->host_lock, flags);

	esp->eh_reset = &eh_reset;

	/* XXX This is too simple... We should add lots of
	 * XXX checks here so that if we find that the chip is
	 * XXX very wedged we return failure immediately so
	 * XXX that we can perform a full chip reset.
	 */
	esp->flags |= ESP_FLAG_RESETTING;
	scsi_esp_cmd(esp, ESP_CMD_RS);

	spin_unlock_irqrestore(esp->host->host_lock, flags);

	ssleep(esp_bus_reset_settle);

	if (!wait_for_completion_timeout(&eh_reset, 5 * HZ)) {
		spin_lock_irqsave(esp->host->host_lock, flags);
		esp->eh_reset = NULL;
		spin_unlock_irqrestore(esp->host->host_lock, flags);

		return FAILED;
	}

	return SUCCESS;
}

/* All bets are off, reset the entire device.  */
static int esp_eh_host_reset_handler(struct scsi_cmnd *cmd)
{
	struct esp *esp = shost_priv(cmd->device->host);
	unsigned long flags;

	spin_lock_irqsave(esp->host->host_lock, flags);
	esp_bootup_reset(esp);
	esp_reset_cleanup(esp);
	spin_unlock_irqrestore(esp->host->host_lock, flags);

	ssleep(esp_bus_reset_settle);

	return SUCCESS;
}

static const char *esp_info(struct Scsi_Host *host)
{
	return "esp";
}

struct scsi_host_template scsi_esp_template = {
	.module			= THIS_MODULE,
	.name			= "esp",
	.info			= esp_info,
	.queuecommand		= esp_queuecommand,
	.target_alloc		= esp_target_alloc,
	.target_destroy		= esp_target_destroy,
	.slave_alloc		= esp_slave_alloc,
	.slave_configure	= esp_slave_configure,
	.slave_destroy		= esp_slave_destroy,
	.eh_abort_handler	= esp_eh_abort_handler,
	.eh_bus_reset_handler	= esp_eh_bus_reset_handler,
	.eh_host_reset_handler	= esp_eh_host_reset_handler,
	.can_queue		= 7,
	.this_id		= 7,
	.sg_tablesize		= SG_ALL,
	.use_clustering		= ENABLE_CLUSTERING,
	.max_sectors		= 0xffff,
	.skip_settle_delay	= 1,
};
EXPORT_SYMBOL(scsi_esp_template);

static void esp_get_signalling(struct Scsi_Host *host)
{
	struct esp *esp = shost_priv(host);
	enum spi_signal_type type;

	if (esp->flags & ESP_FLAG_DIFFERENTIAL)
		type = SPI_SIGNAL_HVD;
	else
		type = SPI_SIGNAL_SE;

	spi_signalling(host) = type;
}

static void esp_set_offset(struct scsi_target *target, int offset)
{
	struct Scsi_Host *host = dev_to_shost(target->dev.parent);
	struct esp *esp = shost_priv(host);
	struct esp_target_data *tp = &esp->target[target->id];

	if (esp->flags & ESP_FLAG_DISABLE_SYNC)
		tp->nego_goal_offset = 0;
	else
		tp->nego_goal_offset = offset;
	tp->flags |= ESP_TGT_CHECK_NEGO;
}

static void esp_set_period(struct scsi_target *target, int period)
{
	struct Scsi_Host *host = dev_to_shost(target->dev.parent);
	struct esp *esp = shost_priv(host);
	struct esp_target_data *tp = &esp->target[target->id];

	tp->nego_goal_period = period;
	tp->flags |= ESP_TGT_CHECK_NEGO;
}

static void esp_set_width(struct scsi_target *target, int width)
{
	struct Scsi_Host *host = dev_to_shost(target->dev.parent);
	struct esp *esp = shost_priv(host);
	struct esp_target_data *tp = &esp->target[target->id];

	tp->nego_goal_width = (width ? 1 : 0);
	tp->flags |= ESP_TGT_CHECK_NEGO;
}

static struct spi_function_template esp_transport_ops = {
	.set_offset		= esp_set_offset,
	.show_offset		= 1,
	.set_period		= esp_set_period,
	.show_period		= 1,
	.set_width		= esp_set_width,
	.show_width		= 1,
	.get_signalling		= esp_get_signalling,
};

static int __init esp_init(void)
{
	BUILD_BUG_ON(sizeof(struct scsi_pointer) <
		     sizeof(struct esp_cmd_priv));

	esp_transport_template = spi_attach_transport(&esp_transport_ops);
	if (!esp_transport_template)
		return -ENODEV;

	return 0;
}

static void __exit esp_exit(void)
{
	spi_release_transport(esp_transport_template);
}

MODULE_DESCRIPTION("ESP SCSI driver core");
MODULE_AUTHOR("David S. Miller (davem@davemloft.net)");
MODULE_LICENSE("GPL");
MODULE_VERSION(DRV_VERSION);

module_param(esp_bus_reset_settle, int, 0);
MODULE_PARM_DESC(esp_bus_reset_settle,
		 "ESP scsi bus reset delay in seconds");

module_param(esp_debug, int, 0);
MODULE_PARM_DESC(esp_debug,
"ESP bitmapped debugging message enable value:\n"
"	0x00000001	Log interrupt events\n"
"	0x00000002	Log scsi commands\n"
"	0x00000004	Log resets\n"
"	0x00000008	Log message in events\n"
"	0x00000010	Log message out events\n"
"	0x00000020	Log command completion\n"
"	0x00000040	Log disconnects\n"
"	0x00000080	Log data start\n"
"	0x00000100	Log data done\n"
"	0x00000200	Log reconnects\n"
"	0x00000400	Log auto-sense data\n"
);

module_init(esp_init);
module_exit(esp_exit);
generated by cgit v1.2.2 (git 2.25.0) at 2025-09-07 17:44:01 -0400