> -----Original Message----- > From: James Bottomley [mailto:James.Bottomley@xxxxxxxxxxxxxxxxxxxxx] > Sent: Thursday, October 1, 2015 1:52 PM > To: KY Srinivasan <kys@xxxxxxxxxxxxx> > Cc: Olaf Hering <olaf@xxxxxxxxx>; Haiyang Zhang > <haiyangz@xxxxxxxxxxxxx>; linux-kernel@xxxxxxxxxxxxxxx; linux- > scsi@xxxxxxxxxxxxxxx > Subject: Re: [PATCH] storvsc: use small sg_tablesize on x86 > > On Thu, 2015-10-01 at 20:30 +0000, KY Srinivasan wrote: > > > > > -----Original Message----- > > > From: Olaf Hering [mailto:olaf@xxxxxxxxx] > > > Sent: Tuesday, September 8, 2015 3:07 AM > > > To: KY Srinivasan <kys@xxxxxxxxxxxxx>; Haiyang Zhang > > > <haiyangz@xxxxxxxxxxxxx>; JBottomley@xxxxxxxx > > > Cc: linux-kernel@xxxxxxxxxxxxxxx; linux-scsi@xxxxxxxxxxxxxxx; Olaf Hering > > > <olaf@xxxxxxxxx> > > > Subject: [PATCH] storvsc: use small sg_tablesize on x86 > > > > > > Reducing the sg_tablesize allows booting of 32bit kernels in VMs, after > > > commit be0cf6ca301c61458dc4aa1a37acf4f58d2ed3d6 ("scsi: storvsc: Set > the > > > tablesize based on the information given by the host") > > > > > > [ 5.567138] hv_storvsc vmbus_1: adjusting sg_tablesize 0x800 -> 0x20 > > > > > > Signed-off-by: Olaf Hering <olaf@xxxxxxxxx> > > Reviewed-by: K. Y. Srinivasan <kys@xxxxxxxxxxxxx> > > > --- > > > drivers/scsi/storvsc_drv.c | 5 +++++ > > > 1 file changed, 5 insertions(+) > > > > > > diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c > > > index 40c43ae..36bf183 100644 > > > --- a/drivers/scsi/storvsc_drv.c > > > +++ b/drivers/scsi/storvsc_drv.c > > > @@ -1907,6 +1907,11 @@ static int storvsc_probe(struct hv_device > *device, > > > * from the host. > > > */ > > > host->sg_tablesize = (stor_device->max_transfer_bytes >> > > > PAGE_SHIFT); > > > +#if defined(CONFIG_X86_32) > > > + dev_warn(&device->device, "adjusting sg_tablesize 0x%x -> 0x%x", > > > + host->sg_tablesize, > > > MAX_MULTIPAGE_BUFFER_COUNT); > > > + host->sg_tablesize = MAX_MULTIPAGE_BUFFER_COUNT; > > > +#endif > > Um, this is a bit architecture specific (I know Azure is x86, but > still). Can you make the define check CONFIG_32BIT rather than > CONFIG_X86_32? Thanks, will do. K. Y > > James > > ��.n��������+%������w��{.n�����{������ܨ}���Ơz�j:+v�����w����ޙ��&�)ߡ�a����z�ޗ���ݢj��w�f
