Aren
/
NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
NetBSD/sys/dev/qbus/rl.c

747 lines
18 KiB
C
Raw Blame History

/* $NetBSD: rl.c,v 1.30 2006/03/29 18:17:36 thorpej Exp $ */
/*
* Copyright (c) 2000 Ludd, University of Lule}, Sweden. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed at Ludd, University of
* Lule}, Sweden and its contributors.
* 4. The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/*
* RL11/RLV11/RLV12 disk controller driver and
* RL01/RL02 disk device driver.
*
* TODO:
* Handle disk errors more gracefully
* Do overlapping seeks on multiple drives
*
* Implementation comments:
*
*/
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: rl.c,v 1.30 2006/03/29 18:17:36 thorpej Exp $");
#include <sys/param.h>
#include <sys/device.h>
#include <sys/systm.h>
#include <sys/conf.h>
#include <sys/disk.h>
#include <sys/disklabel.h>
#include <sys/buf.h>
#include <sys/bufq.h>
#include <sys/stat.h>
#include <sys/dkio.h>
#include <sys/fcntl.h>
#include <sys/event.h>
#include <ufs/ufs/dinode.h>
#include <ufs/ffs/fs.h>
#include <machine/bus.h>
#include <dev/qbus/ubavar.h>
#include <dev/qbus/rlreg.h>
#include <dev/qbus/rlvar.h>
#include "ioconf.h"
#include "locators.h"
static int rlcmatch(struct device *, struct cfdata *, void *);
static void rlcattach(struct device *, struct device *, void *);
static int rlcprint(void *, const char *);
static void rlcintr(void *);
static int rlmatch(struct device *, struct cfdata *, void *);
static void rlattach(struct device *, struct device *, void *);
static void rlcstart(struct rlc_softc *, struct buf *);
static void waitcrdy(struct rlc_softc *);
static void rlcreset(struct device *);
CFATTACH_DECL(rlc, sizeof(struct rlc_softc),
rlcmatch, rlcattach, NULL, NULL);
CFATTACH_DECL(rl, sizeof(struct rl_softc),
rlmatch, rlattach, NULL, NULL);
dev_type_open(rlopen);
dev_type_close(rlclose);
dev_type_read(rlread);
dev_type_write(rlwrite);
dev_type_ioctl(rlioctl);
dev_type_strategy(rlstrategy);
dev_type_dump(rldump);
dev_type_size(rlsize);
const struct bdevsw rl_bdevsw = {
rlopen, rlclose, rlstrategy, rlioctl, rldump, rlsize, D_DISK
};
const struct cdevsw rl_cdevsw = {
rlopen, rlclose, rlread, rlwrite, rlioctl,
nostop, notty, nopoll, nommap, nokqfilter, D_DISK
};
#define MAXRLXFER (RL_BPS * RL_SPT)
#define RL_WREG(reg, val) \
bus_space_write_2(sc->sc_iot, sc->sc_ioh, (reg), (val))
#define RL_RREG(reg) \
bus_space_read_2(sc->sc_iot, sc->sc_ioh, (reg))
static const char *rlstates[] = {
"drive not loaded",
"drive spinning up",
"drive brushes out",
"drive loading heads",
"drive seeking",
"drive ready",
"drive unloading heads",
"drive spun down",
};
static struct dkdriver rldkdriver = {
rlstrategy, minphys
};
static const char *
rlstate(struct rlc_softc *sc, int unit)
{
int i = 0;
do {
RL_WREG(RL_DA, RLDA_GS);
RL_WREG(RL_CS, RLCS_GS|(unit << RLCS_USHFT));
waitcrdy(sc);
} while (((RL_RREG(RL_CS) & RLCS_ERR) != 0) && i++ < 10);
if (i == 10)
return NULL;
i = RL_RREG(RL_MP) & RLMP_STATUS;
return rlstates[i];
}
void
waitcrdy(struct rlc_softc *sc)
{
int i;
for (i = 0; i < 1000; i++) {
DELAY(10000);
if (RL_RREG(RL_CS) & RLCS_CRDY)
return;
}
printf("%s: never got ready\n", sc->sc_dev.dv_xname); /* ?panic? */
}
int
rlcprint(void *aux, const char *name)
{
struct rlc_attach_args *ra = aux;
if (name)
aprint_normal("RL0%d at %s",
ra->type & RLMP_DT ? '2' : '1', name);
aprint_normal(" drive %d", ra->hwid);
return UNCONF;
}
/*
* Force the controller to interrupt.
*/
int
rlcmatch(struct device *parent, struct cfdata *cf, void *aux)
{
struct uba_attach_args *ua = aux;
struct rlc_softc ssc, *sc = &ssc;
int i;
sc->sc_iot = ua->ua_iot;
sc->sc_ioh = ua->ua_ioh;
/* Force interrupt by issuing a "Get Status" command */
RL_WREG(RL_DA, RLDA_GS);
RL_WREG(RL_CS, RLCS_GS|RLCS_IE);
for (i = 0; i < 100; i++) {
DELAY(100000);
if (RL_RREG(RL_CS) & RLCS_CRDY)
return 1;
}
return 0;
}
void
rlcattach(struct device *parent, struct device *self, void *aux)
{
struct rlc_softc *sc = device_private(self);
struct uba_attach_args *ua = aux;
struct rlc_attach_args ra;
int i, error;
sc->sc_iot = ua->ua_iot;
sc->sc_ioh = ua->ua_ioh;
sc->sc_dmat = ua->ua_dmat;
uba_intr_establish(ua->ua_icookie, ua->ua_cvec,
rlcintr, sc, &sc->sc_intrcnt);
evcnt_attach_dynamic(&sc->sc_intrcnt, EVCNT_TYPE_INTR, ua->ua_evcnt,
sc->sc_dev.dv_xname, "intr");
uba_reset_establish(rlcreset, self);
printf("\n");
/*
* The RL11 can only have one transfer going at a time,
* and max transfer size is one track, so only one dmamap
* is needed.
*/
error = bus_dmamap_create(sc->sc_dmat, MAXRLXFER, 1, MAXRLXFER, 0,
BUS_DMA_ALLOCNOW, &sc->sc_dmam);
if (error) {
printf(": Failed to allocate DMA map, error %d\n", error);
return;
}
bufq_alloc(&sc->sc_q, "disksort", BUFQ_SORT_CYLINDER);
for (i = 0; i < RL_MAXDPC; i++) {
waitcrdy(sc);
RL_WREG(RL_DA, RLDA_GS|RLDA_RST);
RL_WREG(RL_CS, RLCS_GS|(i << RLCS_USHFT));
waitcrdy(sc);
ra.type = RL_RREG(RL_MP);
ra.hwid = i;
if ((RL_RREG(RL_CS) & RLCS_ERR) == 0)
config_found(&sc->sc_dev, &ra, rlcprint);
}
}
int
rlmatch(struct device *parent, struct cfdata *cf, void *aux)
{
struct rlc_attach_args *ra = aux;
if (cf->cf_loc[RLCCF_DRIVE] != RLCCF_DRIVE_DEFAULT &&
cf->cf_loc[RLCCF_DRIVE] != ra->hwid)
return 0;
return 1;
}
void
rlattach(struct device *parent, struct device *self, void *aux)
{
struct rl_softc *rc = device_private(self);
struct rlc_attach_args *ra = aux;
struct disklabel *dl;
rc->rc_hwid = ra->hwid;
rc->rc_disk.dk_name = rc->rc_dev.dv_xname;
rc->rc_disk.dk_driver = &rldkdriver;
disk_attach(&rc->rc_disk);
dl = rc->rc_disk.dk_label;
dl->d_npartitions = 3;
strcpy(dl->d_typename, "RL01");
if (ra->type & RLMP_DT)
dl->d_typename[3] = '2';
dl->d_secsize = DEV_BSIZE; /* XXX - wrong, but OK for now */
dl->d_nsectors = RL_SPT/2;
dl->d_ntracks = RL_SPD;
dl->d_ncylinders = ra->type & RLMP_DT ? RL_TPS02 : RL_TPS01;
dl->d_secpercyl = dl->d_nsectors * dl->d_ntracks;
dl->d_secperunit = dl->d_ncylinders * dl->d_secpercyl;
dl->d_partitions[0].p_size = dl->d_partitions[2].p_size =
dl->d_secperunit;
dl->d_partitions[0].p_offset = dl->d_partitions[2].p_offset = 0;
dl->d_interleave = dl->d_headswitch = 1;
dl->d_bbsize = BBSIZE;
dl->d_sbsize = SBLOCKSIZE;
dl->d_rpm = 2400;
dl->d_type = DTYPE_DEC;
printf(": %s, %s\n", dl->d_typename,
rlstate((struct rlc_softc *)parent, ra->hwid));
/*
* XXX We should try to discovery wedges here, but
* XXX that would mean loading up the pack and being
* XXX able to do I/O. Should use config_defer() here.
*/
}
int
rlopen(dev_t dev, int flag, int fmt, struct lwp *l)
{
int error, part, unit, mask;
struct disklabel *dl;
struct rlc_softc *sc;
struct rl_softc *rc;
const char *msg;
/*
* Make sure this is a reasonable open request.
*/
unit = DISKUNIT(dev);
if (unit >= rl_cd.cd_ndevs)
return ENXIO;
rc = rl_cd.cd_devs[unit];
if (rc == 0)
return ENXIO;
part = DISKPART(dev);
if ((error = lockmgr(&rc->rc_disk.dk_openlock, LK_EXCLUSIVE,
NULL)) != 0)
return (error);
/*
* If there are wedges, and this is not RAW_PART, then we
* need to fail.
*/
if (rc->rc_disk.dk_nwedges != 0 && part != RAW_PART) {
error = EBUSY;
goto bad1;
}
sc = (struct rlc_softc *)device_parent(&rc->rc_dev);
/* Check that the disk actually is useable */
msg = rlstate(sc, rc->rc_hwid);
if (msg == NULL || msg == rlstates[RLMP_UNLOAD] ||
msg == rlstates[RLMP_SPUNDOWN]) {
error = ENXIO;
goto bad1;
}
/*
* If this is the first open; read in where on the disk we are.
*/
dl = rc->rc_disk.dk_label;
if (rc->rc_state == DK_CLOSED) {
u_int16_t mp;
int maj;
RL_WREG(RL_CS, RLCS_RHDR|(rc->rc_hwid << RLCS_USHFT));
waitcrdy(sc);
mp = RL_RREG(RL_MP);
rc->rc_head = ((mp & RLMP_HS) == RLMP_HS);
rc->rc_cyl = (mp >> 7) & 0777;
rc->rc_state = DK_OPEN;
/* Get disk label */
printf("%s: ", rc->rc_dev.dv_xname);
maj = cdevsw_lookup_major(&rl_cdevsw);
if ((msg = readdisklabel(MAKEDISKDEV(maj,
device_unit(&rc->rc_dev), RAW_PART), rlstrategy, dl, NULL)))
printf("%s: ", msg);
printf("size %d sectors\n", dl->d_secperunit);
}
if (part >= dl->d_npartitions) {
error = ENXIO;
goto bad1;
}
mask = 1 << part;
switch (fmt) {
case S_IFCHR:
rc->rc_disk.dk_copenmask |= mask;
break;
case S_IFBLK:
rc->rc_disk.dk_bopenmask |= mask;
break;
}
rc->rc_disk.dk_openmask |= mask;
(void) lockmgr(&rc->rc_disk.dk_openlock, LK_RELEASE, NULL);
return 0;
bad1:
(void) lockmgr(&rc->rc_disk.dk_openlock, LK_RELEASE, NULL);
return (error);
}
int
rlclose(dev_t dev, int flag, int fmt, struct lwp *l)
{
int error, unit = DISKUNIT(dev);
struct rl_softc *rc = rl_cd.cd_devs[unit];
int mask = (1 << DISKPART(dev));
if ((error = lockmgr(&rc->rc_disk.dk_openlock, LK_EXCLUSIVE,
NULL)) != 0)
return (error);
switch (fmt) {
case S_IFCHR:
rc->rc_disk.dk_copenmask &= ~mask;
break;
case S_IFBLK:
rc->rc_disk.dk_bopenmask &= ~mask;
break;
}
rc->rc_disk.dk_openmask =
rc->rc_disk.dk_copenmask | rc->rc_disk.dk_bopenmask;
if (rc->rc_disk.dk_openmask == 0)
rc->rc_state = DK_CLOSED; /* May change pack */
(void) lockmgr(&rc->rc_disk.dk_openlock, LK_RELEASE, NULL);
return 0;
}
void
rlstrategy(struct buf *bp)
{
struct disklabel *lp;
struct rlc_softc *sc;
struct rl_softc *rc;
int unit, s, err;
/*
* Make sure this is a reasonable drive to use.
*/
unit = DISKUNIT(bp->b_dev);
if (unit > rl_cd.cd_ndevs || (rc = rl_cd.cd_devs[unit]) == NULL) {
bp->b_error = ENXIO;
bp->b_flags |= B_ERROR;
goto done;
}
if (rc->rc_state != DK_OPEN) /* How did we end up here at all? */
panic("rlstrategy: state impossible");
lp = rc->rc_disk.dk_label;
if ((err = bounds_check_with_label(&rc->rc_disk, bp, 1)) <= 0)
goto done;
if (bp->b_bcount == 0)
goto done;
bp->b_rawblkno =
bp->b_blkno + lp->d_partitions[DISKPART(bp->b_dev)].p_offset;
bp->b_cylinder = bp->b_rawblkno / lp->d_secpercyl;
sc = (struct rlc_softc *)device_parent(&rc->rc_dev);
s = splbio();
BUFQ_PUT(sc->sc_q, bp);
rlcstart(sc, 0);
splx(s);
return;
done: biodone(bp);
}
int
rlioctl(dev_t dev, u_long cmd, caddr_t addr, int flag, struct lwp *l)
{
struct rl_softc *rc = rl_cd.cd_devs[DISKUNIT(dev)];
struct disklabel *lp = rc->rc_disk.dk_label;
int err = 0;
#ifdef __HAVE_OLD_DISKLABEL
struct disklabel newlabel;
#endif
switch (cmd) {
case DIOCGDINFO:
bcopy(lp, addr, sizeof (struct disklabel));
break;
#ifdef __HAVE_OLD_DISKLABEL
case ODIOCGDINFO:
newlabel = *lp;
if (newlabel.d_npartitions > OLDMAXPARTITIONS)
return ENOTTY;
bcopy(&newlabel, addr, sizeof (struct olddisklabel));
break;
#endif
case DIOCGPART:
((struct partinfo *)addr)->disklab = lp;
((struct partinfo *)addr)->part =
&lp->d_partitions[DISKPART(dev)];
break;
case DIOCSDINFO:
case DIOCWDINFO:
#ifdef __HAVE_OLD_DISKLABEL
case ODIOCWDINFO:
case ODIOCSDINFO:
#endif
{
struct disklabel *tp;
#ifdef __HAVE_OLD_DISKLABEL
if (cmd == ODIOCSDINFO || cmd == ODIOCWDINFO) {
memset(&newlabel, 0, sizeof newlabel);
memcpy(&newlabel, addr, sizeof (struct olddisklabel));
tp = &newlabel;
} else
#endif
tp = (struct disklabel *)addr;
if ((flag & FWRITE) == 0)
err = EBADF;
else {
if ((err = lockmgr(&rc->rc_disk.dk_openlock,
LK_EXCLUSIVE, NULL)) != 0)
break;
err = ((
#ifdef __HAVE_OLD_DISKLABEL
cmd == ODIOCSDINFO ||
#endif
cmd == DIOCSDINFO) ?
setdisklabel(lp, tp, 0, 0) :
writedisklabel(dev, rlstrategy, lp, 0));
(void) lockmgr(&rc->rc_disk.dk_openlock,
LK_RELEASE, NULL);
}
break;
}
case DIOCWLABEL:
if ((flag & FWRITE) == 0)
err = EBADF;
break;
case DIOCAWEDGE:
{
struct dkwedge_info *dkw = (void *) addr;
if ((flag & FWRITE) == 0)
return (EBADF);
/* If the ioctl happens here, the parent is us. */
strcpy(dkw->dkw_parent, rc->rc_dev.dv_xname);
return (dkwedge_add(dkw));
}
case DIOCDWEDGE:
{
struct dkwedge_info *dkw = (void *) addr;
if ((flag & FWRITE) == 0)
return (EBADF);
/* If the ioctl happens here, the parent is us. */
strcpy(dkw->dkw_parent, rc->rc_dev.dv_xname);
return (dkwedge_del(dkw));
}
case DIOCLWEDGES:
{
struct dkwedge_list *dkwl = (void *) addr;
return (dkwedge_list(&rc->rc_disk, dkwl, l));
}
default:
err = ENOTTY;
}
return err;
}
int
rlsize(dev_t dev)
{
struct disklabel *dl;
struct rl_softc *rc;
int size, unit = DISKUNIT(dev);
if ((unit >= rl_cd.cd_ndevs) || ((rc = rl_cd.cd_devs[unit]) == 0))
return -1;
dl = rc->rc_disk.dk_label;
size = dl->d_partitions[DISKPART(dev)].p_size *
(dl->d_secsize / DEV_BSIZE);
return size;
}
int
rldump(dev_t dev, daddr_t blkno, caddr_t va, size_t size)
{
/* Not likely... */
return 0;
}
int
rlread(dev_t dev, struct uio *uio, int ioflag)
{
return (physio(rlstrategy, NULL, dev, B_READ, minphys, uio));
}
int
rlwrite(dev_t dev, struct uio *uio, int ioflag)
{
return (physio(rlstrategy, NULL, dev, B_WRITE, minphys, uio));
}
static const char *rlerr[] = {
"no",
"operation incomplete",
"read data CRC",
"header CRC",
"data late",
"header not found",
"",
"",
"non-existent memory",
"memory parity error",
"",
"",
"",
"",
"",
"",
};
void
rlcintr(void *arg)
{
struct rlc_softc *sc = arg;
struct buf *bp;
u_int16_t cs;
bp = sc->sc_active;
if (bp == 0) {
printf("%s: strange interrupt\n", sc->sc_dev.dv_xname);
return;
}
bus_dmamap_unload(sc->sc_dmat, sc->sc_dmam);
sc->sc_active = 0;
cs = RL_RREG(RL_CS);
if (cs & RLCS_ERR) {
int error = (cs & RLCS_ERRMSK) >> 10;
printf("%s: %s\n", sc->sc_dev.dv_xname, rlerr[error]);
bp->b_flags |= B_ERROR;
bp->b_error = EIO;
bp->b_resid = bp->b_bcount;
sc->sc_bytecnt = 0;
}
if (sc->sc_bytecnt == 0) /* Finished transfer */
biodone(bp);
rlcstart(sc, sc->sc_bytecnt ? bp : 0);
}
/*
* Start routine. First position the disk to the given position,
* then start reading/writing. An optimization would be to be able
* to handle overlapping seeks between disks.
*/
void
rlcstart(struct rlc_softc *sc, struct buf *ob)
{
struct disklabel *lp;
struct rl_softc *rc;
struct buf *bp;
int bn, cn, sn, tn, blks, err;
if (sc->sc_active)
return; /* Already doing something */
if (ob == 0) {
bp = BUFQ_GET(sc->sc_q);
if (bp == NULL)
return; /* Nothing to do */
sc->sc_bufaddr = bp->b_data;
sc->sc_diskblk = bp->b_rawblkno;
sc->sc_bytecnt = bp->b_bcount;
bp->b_resid = 0;
} else
bp = ob;
sc->sc_active = bp;
rc = rl_cd.cd_devs[DISKUNIT(bp->b_dev)];
bn = sc->sc_diskblk;
lp = rc->rc_disk.dk_label;
if (bn) {
cn = bn / lp->d_secpercyl;
sn = bn % lp->d_secpercyl;
tn = sn / lp->d_nsectors;
sn = sn % lp->d_nsectors;
} else
cn = sn = tn = 0;
/*
* Check if we have to position disk first.
*/
if (rc->rc_cyl != cn || rc->rc_head != tn) {
u_int16_t da = RLDA_SEEK;
if (cn > rc->rc_cyl)
da |= ((cn - rc->rc_cyl) << RLDA_CYLSHFT) | RLDA_DIR;
else
da |= ((rc->rc_cyl - cn) << RLDA_CYLSHFT);
if (tn)
da |= RLDA_HSSEEK;
waitcrdy(sc);
RL_WREG(RL_DA, da);
RL_WREG(RL_CS, RLCS_SEEK | (rc->rc_hwid << RLCS_USHFT));
waitcrdy(sc);
rc->rc_cyl = cn;
rc->rc_head = tn;
}
RL_WREG(RL_DA, (cn << RLDA_CYLSHFT) | (tn ? RLDA_HSRW : 0) | (sn << 1));
blks = sc->sc_bytecnt/DEV_BSIZE;
if (sn + blks > RL_SPT/2)
blks = RL_SPT/2 - sn;
RL_WREG(RL_MP, -(blks*DEV_BSIZE)/2);
err = bus_dmamap_load(sc->sc_dmat, sc->sc_dmam, sc->sc_bufaddr,
(blks*DEV_BSIZE), (bp->b_flags & B_PHYS ? bp->b_proc : 0),
BUS_DMA_NOWAIT);
if (err)
panic("%s: bus_dmamap_load failed: %d",
sc->sc_dev.dv_xname, err);
RL_WREG(RL_BA, (sc->sc_dmam->dm_segs[0].ds_addr & 0xffff));
/* Count up vars */
sc->sc_bufaddr += (blks*DEV_BSIZE);
sc->sc_diskblk += blks;
sc->sc_bytecnt -= (blks*DEV_BSIZE);
if (bp->b_flags & B_READ)
RL_WREG(RL_CS, RLCS_IE|RLCS_RD|(rc->rc_hwid << RLCS_USHFT));
else
RL_WREG(RL_CS, RLCS_IE|RLCS_WD|(rc->rc_hwid << RLCS_USHFT));
}
/*
* Called once per controller when an ubareset occurs.
* Retracts all disks and restarts active transfers.
*/
void
rlcreset(struct device *dev)
{
struct rlc_softc *sc = (struct rlc_softc *)dev;
struct rl_softc *rc;
int i;
u_int16_t mp;
for (i = 0; i < rl_cd.cd_ndevs; i++) {
if ((rc = rl_cd.cd_devs[i]) == NULL)
continue;
if (rc->rc_state != DK_OPEN)
continue;
printf(" %s", rc->rc_dev.dv_xname);
RL_WREG(RL_CS, RLCS_RHDR|(rc->rc_hwid << RLCS_USHFT));
waitcrdy(sc);
mp = RL_RREG(RL_MP);
rc->rc_head = ((mp & RLMP_HS) == RLMP_HS);
rc->rc_cyl = (mp >> 7) & 0777;
}
if (sc->sc_active == 0)
return;
BUFQ_PUT(sc->sc_q, sc->sc_active);
sc->sc_active = 0;
rlcstart(sc, 0);
}
Reference in New Issue View Git Blame Copy Permalink