SCSI problem solved
Thomas Bogendoerfer
maintainerRight now I'm doing some stress test with my Indy (dd from DAT and untaring a tar file from one disk to another at the same time). And it hasn't crashed. Below is the patch I'm using. If it works for others too, I'll check it in.
Thomas.
Index: sgiwd93.c
===================================================================
RCS file: /var/mips/linus/cvs/linux/drivers/scsi/sgiwd93.c,v retrieving revision 1.8
sgiwd93.c +9 -9
@@ -77,7 +77,7 @@static int dma_setup (Scsi_Cmnd *cmd,int datainp) { -struct WD33C93_hostdata *hdata =CMDHOSTDATA (cmd); +struct WD33C93_hostdata *hdata = (struct WD33C93_hostdata *)cmd->host->hostdata;wd33c93_regs *regp = hdata->regp;struct hpc3_scsiregs *hregs = (struct hpc3_scsiregs *) cmd->host->base;struct hpc_chunk *hcp = (struct hpc_chunk *) hdata->dma_bounce_buffer;
@@ -89,14 +89,14 @@ hdata->dma_dir = datainp; -if (cmd->use_sg) { +if (cmd->SCp.buffers_residual) {struct scatterlist *slp = cmd->SCp.buffer;int i, totlen =0 ;#ifdef DEBUG_DMA printk ("SCLIST<" );#endif -for (i =0 ; i <= (cmd->use_sg -1 ); i++, hcp++) { +for (i =0 ; i <= cmd->SCp.buffers_residual; i++, hcp++) {#ifdef DEBUG_DMA printk ("[%p,%d]" , slp[i].address, slp[i].length);#endif
@@ -146,7 +146,7 @@static void dma_stop (struct Scsi_Host *instance,Scsi_Cmnd *SCpnt,int status) { -struct WD33C93_hostdata *hdata =INSTHOSTDATA (instance); +struct WD33C93_hostdata *hdata = (struct WD33C93_hostdata *)instance->hostdata;wd33c93_regs *regp = hdata->regp;struct hpc3_scsiregs *hregs = (struct hpc3_scsiregs *) SCpnt->host->base;
@@ -163,7 +163,7 @@ hregs->ctrl =0 ;/* See how far we got and update scatterlist state if necessary. */ -if (SCpnt->use_sg) { +if (SCpnt->SCp.buffers_residual) {struct scatterlist *slp = SCpnt->SCp.buffer;int totlen, wd93_residual, transferred, i;
@@ -183,7 +183,7 @@#ifdef DEBUG_DMA printk ("Jed was here..." );#endif -for (i =0 ; i <= (SCpnt->use_sg -1 ); i++) { +for (i =0 ; i <= SCpnt->SCp.buffers_residual; i++) {if (slp[i].length >= transferred)break ; transferred -= slp[i].length;
@@ -193,10 +193,10 @@#ifdef DEBUG_DMA printk ("did it all..." );#endif - i = (SCpnt->use_sg -1 ); + i = SCpnt->SCp.buffers_residual; } SCpnt->SCp.buffer = &slp[i]; - SCpnt->SCp.buffers_residual = (SCpnt->use_sg -1 - i); + SCpnt->SCp.buffers_residual = SCpnt->SCp.buffers_residual - i; SCpnt->SCp.ptr = (char *) slp[i].address; SCpnt->SCp.this_residual = slp[i].length; }
@@ -244,7 +244,7 @@wd33c93_init (sgiwd93_host, (wd33c93_regs *)0xbfbc0003 , dma_setup, dma_stop,WD33C93_FS_16_20 ); - hdata =INSTHOSTDATA (sgiwd93_host); + hdata = (struct WD33C93_hostdata *)sgiwd93_host->hostdata; hdata->no_sync =0 ; hdata->dma_bounce_buffer = (uchar *) (KSEG1ADDR (buf));dma_cache_wback_inv ((unsigned long ) buf,PAGE_SIZE );
Signature · 3 lines
See, you not only have to be a good coder to create a system like Linux,
you have to be a sneaky bastard too ;-)
[Linus Torvalds in <[email protected]>]Quoted 3 lines · Thomas Bogendoerfer
Right now I'm doing some stress test with my Indy (dd from DAT and untaring a tar file from one disk to another at the same time). And it hasn't crashed. Below is the patch I'm using. If it works for others too, I'll check it in.
Side effect: lmbench says disk write bandwidth has roughly trippled.
Ralf