1Sat Dec 30 21:30 2000 Gerard Roudier 2 * version sym-2.1.0-20001230 3 - Initial release of SYM-2. 4 5Mon Jan 08 21:30 2001 Gerard Roudier 6 * version sym-2.1.1-20010108 7 - Change a couple of defines containing ncr or NCR by their 8 equivalent containing sym or SYM instead. 9 10Sun Jan 14 22:30 2001 Gerard Roudier 11 * version sym-2.1.2-20010114 12 - Fix a couple of printfs: 13 * Add the target number to the display of transfer parameters. 14 * Make the display of TCQ and queue depth clearer. 15 16Wed Jan 17 23:30 2001 Gerard Roudier 17 * version sym-2.1.3-20010117 18 - Wrong residual values were returned in some situations. 19 This broke cdrecord with linux-2.4.0, for example. 20 21Sat Jan 20 18:00 2001 Gerard Roudier 22 * version sym-2.1.4-20010120 23 - Add year 2001 to Copyright. 24 - A tiny bug in the dma memory freeing path has been fixed. 25 (Driver unload failed with a bad address reference). 26 27Wed Jan 24 21:00 2001 Gerard Roudier 28 * version sym-2.1.5-20010124 29 - Make the driver work under Linux-2.4.x when statically linked 30 with the kernel. 31 - Check against memory allocation failure for SCRIPTZ and add the 32 missing free of this memory on instance detach. 33 - Check against GPIO3 pulled low for HVD controllers (driver did 34 just the opposite). 35 Misdetection of BUS mode was triggered on module reload only, 36 since BIOS settings were trusted instead on first load. 37 38Wed Feb 7 21:00 2001 Gerard Roudier 39 * version sym-2.1.6-20010207 40 - Call pci_enable_device() as wished by kernel maintainers. 41 - Change the sym_queue_scsiio() interface. 42 This is intended to simplify portability. 43 - Move the code intended to deal with the downloading of SCRIPTS 44 from SCRIPTS :) in the patch method (was wrongly placed in 45 the SCRIPTS setup method). 46 - Add a missing cpu_to_scr() (np->abort_tbl.addr) 47 - Remove a wrong cpu_to_scr() (np->targtbl_ba) 48 - Cleanup a bit the PPR failure recovery code. 49 50Sat Mar 3 21:00 2001 Gerard Roudier 51 - Add option SYM_OPT_ANNOUNCE_TRANSFER_RATE and move the 52 corresponding code to file sym_misc.c. 53 Also move the code that sniffes INQUIRY to sym_misc.c. 54 This allows to share the corresponding code with NetBSD 55 without polluating the core driver source (sym_hipd.c). 56 - Add optional code that handles IO timeouts from the driver. 57 (not used under Linux, but required for NetBSD) 58 - Do not assume any longer that PAGE_SHIFT and PAGE_SIZE are 59 defined at compile time, as at least NetBSD uses variables 60 in memory for that. 61 - Refine a work-around for the C1010-33 that consists in 62 disabling internal LOAD/STORE. Was applied up to revision 1. 63 Is now only applied to revision 0. 64 - Some code reorganisations due to code moves between files. 65 66Tues Apr 10 21:00 2001 Gerard Roudier 67 * version sym-2.1.9-20010412 68 - Reset 53C896 and 53C1010 chip according to the manual. 69 (i.e.: set the ABRT bit in ISTAT if SCRIPTS are running) 70 - Set #LUN in request sense only if scsi version <= 2 and 71 #LUN <= 7. 72 - Set busy_itl in LCB to 1 if the LCB is allocated and a 73 SCSI command is active. This is a simplification. 74 - In sym_hcb_free(), do not scan the free_ccbq if no CCBs 75 has been allocated. This fixes a panic if attach failed. 76 - Add DT/ST (double/simple transition) in the transfer 77 negotiation announce. 78 - Forces the max number of tasks per LUN to at least 64. 79 - Use pci_set_dma_mask() for linux-2.4.3 and above. 80 - A couple of comments fixes. 81 82Wed May 22:00 2001 Gerard Roudier 83 * version sym-2.1.10-20010509 84 - Mask GPCNTL against 0x1c (was 0xfc) for the reading of the NVRAM. 85 This ensure LEDC bit will not be set on 896 and later chips. 86 Fix sent by Chip Salzenberg <chip@perlsupport.com>. 87 - Define the number of PQS BUSes supported. 88 Fix sent by Stig Telfer <stig@api-networks.com> 89 - Miscellaneous common code rearrangements due to NetBSD accel 90 ioctl support, without impact on Linux (hopefully). 91 92Mon July 2 12:00 2001 Gerard Roudier 93 * version sym-2.1.11-20010702 94 - Add Tekram 390 U2B/U2W SCSI LED handling. 95 Submitted by Chip Salzenberg <chip@valinux.com> 96 - Add call to scsi_set_pci_device() for kernels >= 2.4.4. 97 - Check pci dma mapping failures and complete the IO with some 98 error when such mapping fails. 99 - Fill in instance->max_cmd_len for kernels > 2.4.0. 100 - A couple of tiny fixes ... 101 102Sun Sep 9 18:00 2001 Gerard Roudier 103 * version sym-2.1.12-20010909 104 - Change my email address. 105 - Add infrastructure for the forthcoming 64 bit DMA addressing support. 106 (Based on PCI 64 bit patch from David S. Miller) 107 - Do not use anymore vm_offset_t type. 108 109Sat Sep 15 20:00 2001 Gerard Roudier 110 * version sym-2.1.13-20010916 111 - Add support for 64 bit DMA addressing using segment registers. 112 16 registers for up to 4 GB x 16 -> 64 GB. 113 114Sat Sep 22 12:00 2001 Gerard Roudier 115 * version sym-2.1.14-20010922 116 - Complete rewrite of the eh handling. The driver is now using a 117 semaphore in order to behave synchronously as required by the eh 118 threads. A timer is also used to prevent from waiting indefinitely. 119 120Sun Sep 30 17:00 2001 Gerard Roudier 121 * version sym-2.1.15-20010930 122 - Include <linux/module.h> unconditionally as expected by latest 123 kernels. 124 - Use del_timer_sync() for recent kernels to kill the driver timer 125 on module release. 126 127Sun Oct 28 15:00 2001 Gerard Roudier 128 * version sym-2.1.16-20011028 129 - Slightly simplify driver configuration. 130 - Prepare a new patch against linux-2.4.13. 131 132Sat Nov 17 10:00 2001 Gerard Roudier 133 * version sym-2.1.17 134 - Fix a couple of gcc/gcc3 warnings. 135 - Allocate separately from the HCB the array for CCBs hashed by DSA. 136 All driver memory allocations are now not greater than 1 PAGE 137 even on PPC64 / 4KB PAGE surprising setup. 138 139Sat Dec 01 18:00 2001 Gerard Roudier 140 * version sym-2.1.17a 141 - Use u_long instead of U32 for the IO base cookie. This is more 142 consistent with what archs are expecting. 143 - Use MMIO per default for Power PC instead of some fake normal IO, 144 as Paul Mackerras stated that MMIO works fine now on this arch. 145