Ricardo B. Marliere [Mon, 19 Feb 2024 11:45:53 +0000 (08:45 -0300)]
scsi: core: Constify the struct device_type usage
Since commit aed65af1cc2f ("drivers: make device_type const"), the driver
core can properly handle constant struct device_type. Move the
scsi_host_type, scsi_target_type and scsi_dev_type variables to be constant
structures as well, placing it into read-only memory which can not be
modified at runtime.
Justin Stitt [Wed, 21 Feb 2024 23:50:26 +0000 (23:50 +0000)]
scsi: libfc: replace deprecated strncpy() with memcpy()
strncpy() is deprecated [1] and as such we should use different apis to
copy string data.
We can see that ct is NUL-initialized with fc_ct_hdr_fill:
| ct = fc_ct_hdr_fill(fp, op, sizeof(struct fc_ns_rspn) + len,
...
In fc_ct_hdr_fill():
| memset(ct, 0, ct_plen);
We also calculate the length of the source string:
| len = strnlen(fc_host_symbolic_name(lport->host), 255);
...then this argument is used in strncpy(), which is bad because the
pattern of (dest, src, strlen(src)) usually leaves the destination
buffer without NUL-termination. However, it looks as though we do not
require NUL-termination since fr_name is part of a seq_buf-like
structure wherein its length is monitored:
| struct fc_ns_rspn {
| struct fc_ns_fid fr_fid; /* port ID object */
| __u8 fr_name_len;
| char fr_name[];
| } __attribute__((__packed__));
So, this is really just a byte copy into a length-bounded buffer. Let's use
memcpy().
Justin Stitt [Mon, 26 Feb 2024 23:53:44 +0000 (23:53 +0000)]
scsi: lpfc: Replace deprecated strncpy() with strscpy()
strncpy() is deprecated for use on NUL-terminated destination strings [1]
and as such we should prefer more robust and less ambiguous string
interfaces.
We expect ae->value_string to be NUL-terminated because there's a comment
that says as much; these attr strings are also used with other string APIs,
further cementing the fact.
Now, the question of whether or not to NUL-pad the destination buffer:
lpfc_fdmi_rprt_defer() initializes vports (all zero-initialized), then we
call lpfc_fdmi_cmd() with each vport and a mask. Then, inside of
lpfc_fdmi_cmd() we check each bit in the mask to invoke the proper
callback. Importantly, the zero-initialized vport is passed in as the
"attr" parameter. Seeing this:
| struct lpfc_fdmi_attr_string *ae = attr;
... we can tell that ae->value_string is entirely zero-initialized. Due
to this, NUL-padding is _not_ required as it would be redundant.
Considering the above, a suitable replacement is strscpy() [2].
Arnd Bergmann [Thu, 22 Feb 2024 12:44:07 +0000 (13:44 +0100)]
scsi: bfa: Fix function pointer type mismatch for state machines
The bfa driver is full of state machines and a generic abstraction layer
for them. This relies on casting function pointers, but that is no longer
allowed when CONFIG_CFI_CLANG is enabled and causes a huge number of
warnings like:
drivers/scsi/bfa/bfad.c:169:3: error: cast from 'void (*)(struct bfad_s *, enum bfad_sm_event)' to 'bfa_sm_t' (aka 'void (*)(void *, int)') converts to incompatible function type [-Werror,-Wcast-function-type-strict]
bfa_sm_set_state(bfad, bfad_sm_created);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Rework the mechanism to no longer require the function pointer casts, by
having separate types for each individual state machine. This in turn
requires moving the enum definitions for each state machine into the header
files in order to define the typedef.
Arnd Bergmann [Thu, 22 Feb 2024 12:44:06 +0000 (13:44 +0100)]
scsi: bfa: Fix function pointer type mismatch for hcb_qe->cbfn
Some callback functions used here take a boolean argument, others take a
status argument. This breaks KCFI type checking, so clang now warns about
the function pointer cast:
drivers/scsi/bfa/bfad_bsg.c:2138:29: error: cast from 'void (*)(void *, enum bfa_status)' to 'bfa_cb_cbfn_t' (aka 'void (*)(void *, enum bfa_boolean)') converts to incompatible function type [-Werror,-Wcast-function-type-strict]
Assuming the code is actually correct here and the callers always match the
argument types of the callee, rework this to replace the explicit cast with
a union of the two pointer types. This does not change the behavior of the
code, so if something is actually broken here, a larger rework may be
necessary.
Fixes: 37ea0558b87a ("[SCSI] bfa: Added support to collect and reset fcport stats") Fixes: 3ec4f2c8bff2 ("[SCSI] bfa: Added support to configure QOS and collect stats.") Reviewed-by: Kees Cook <keescook@chromium.org> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Link: https://lore.kernel.org/r/20240222124433.2046570-1-arnd@kernel.org Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Arnd Bergmann [Tue, 13 Feb 2024 10:05:00 +0000 (11:05 +0100)]
scsi: csiostor: Avoid function pointer casts
csiostor uses function pointer casts to keep the csio_ln_ev state machine
hidden, but this causes warnings about control flow integrity (KCFI)
violations in clang-16 and higher:
drivers/scsi/csiostor/csio_lnode.c:1098:33: error: cast from 'void (*)(struct csio_lnode *, enum csio_ln_ev)' to 'csio_sm_state_t' (aka 'void (*)(void *, unsigned int)') converts to incompatible function type [-Werror,-Wcast-function-type-strict]
1098 | return (csio_get_state(ln) == ((csio_sm_state_t)csio_lns_ready));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/scsi/csiostor/csio_lnode.c:1369:29: error: cast from 'void (*)(struct csio_lnode *, enum csio_ln_ev)' to 'csio_sm_state_t' (aka 'void (*)(void *, unsigned int)') converts to incompatible function type [-Werror,-Wcast-function-type-strict]
1369 | if (csio_get_state(ln) == ((csio_sm_state_t)csio_lns_uninit)) {
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/scsi/csiostor/csio_lnode.c:1373:29: error: cast from 'void (*)(struct csio_lnode *, enum csio_ln_ev)' to 'csio_sm_state_t' (aka 'void (*)(void *, unsigned int)') converts to incompatible function type [-Werror,-Wcast-function-type-strict]
1373 | if (csio_get_state(ln) == ((csio_sm_state_t)csio_lns_ready)) {
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/scsi/csiostor/csio_lnode.c:1377:29: error: cast from 'void (*)(struct csio_lnode *, enum csio_ln_ev)' to 'csio_sm_state_t' (aka 'void (*)(void *, unsigned int)') converts to incompatible function type [-Werror,-Wcast-function-type-strict]
1377 | if (csio_get_state(ln) == ((csio_sm_state_t)csio_lns_offline)) {
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Move the enum into a shared header so the correct types can be used without
the need for casts.
Ricardo B. Marliere [Mon, 12 Feb 2024 18:38:37 +0000 (15:38 -0300)]
scsi: core: Make scsi_bus_type const
Since commit d492cc2573a0 ("driver core: device.h: make struct bus_type a
const *"), the driver core can properly handle constant struct bus_type,
move the scsi_bus_type variable to be a constant structure as well, placing
it into read-only memory which can not be modified at runtime.
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net> Link: https://lore.kernel.org/r/20240212-bus_cleanup-scsi2-v2-1-65004493ff09@marliere.net Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Lukas Bulwahn [Wed, 7 Feb 2024 14:56:03 +0000 (15:56 +0100)]
scsi: core: Really include kunit tests with SCSI_LIB_KUNIT_TEST
Commit 25a1f7a0a1fe ("scsi: core: Add kunit tests for
scsi_check_passthrough()") adds the config SCSI_LIB_KUNIT_TEST and
corresponding tests. Due to naming confusion, the actual tests would only
be included when the non-existing config SCSI_KUNIT_TEST is enabled (note
this missing 'LIB' in the config name). So, they are basically dead right
now.
Ricardo B. Marliere [Sun, 4 Feb 2024 20:48:26 +0000 (17:48 -0300)]
scsi: target: tcm_loop: Make tcm_loop_lld_bus const
Now that the driver core can properly handle constant struct bus_type, move
the tcm_loop_lld_bus variable to be a constant structure as well, placing
it into read-only memory which can not be modified at runtime.
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net> Link: https://lore.kernel.org/r/20240204-bus_cleanup-target-v1-1-96106936c4ab@marliere.net Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Ricardo B. Marliere [Sat, 3 Feb 2024 18:39:02 +0000 (15:39 -0300)]
scsi: scsi_debug: Make pseudo_lld_bus const
Now that the driver core can properly handle constant struct bus_type, move
the pseudo_lld_bus variable to be a constant structure as well, placing it
into read-only memory which can not be modified at runtime.
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net> Link: https://lore.kernel.org/r/20240203-bus_cleanup-scsi-v1-3-6f552fb24f71@marliere.net Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Ricardo B. Marliere [Sat, 3 Feb 2024 18:39:01 +0000 (15:39 -0300)]
scsi: iscsi: Make iscsi_flashnode_bus const
Now that the driver core can properly handle constant struct bus_type, move
the iscsi_flashnode_bus variable to be a constant structure as well,
placing it into read-only memory which can not be modified at runtime.
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net> Link: https://lore.kernel.org/r/20240203-bus_cleanup-scsi-v1-2-6f552fb24f71@marliere.net Reviewed-by: Lee Duncan <lduncan@suse.com> Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Ricardo B. Marliere [Sat, 3 Feb 2024 18:39:00 +0000 (15:39 -0300)]
scsi: fcoe: Make fcoe_bus_type const
Now that the driver core can properly handle constant struct bus_type, move
the fcoe_bus_type variable to be a constant structure as well, placing it
into read-only memory which can not be modified at runtime.
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Suggested-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Ricardo B. Marliere <ricardo@marliere.net> Link: https://lore.kernel.org/r/20240203-bus_cleanup-scsi-v1-1-6f552fb24f71@marliere.net Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Martin K. Petersen [Tue, 6 Feb 2024 01:52:37 +0000 (20:52 -0500)]
Merge patch series "Update lpfc to revision 14.4.0.0"
Justin Tee <justintee8345@gmail.com> says:
Update lpfc to revision 14.4.0.0
This patch set contains fixes identified by static code analyzers,
updates to log messaging, bug fixes related to discovery and
congestion management, and clean up patches regarding the abuse of
shost lock in the driver.
The patches were cut against Martin's 6.9/scsi-queue tree.
Justin Tee [Wed, 31 Jan 2024 18:51:10 +0000 (10:51 -0800)]
scsi: lpfc: Change lpfc_vport load_flag member into a bitmask
In attempt to reduce the amount of unnecessary shost_lock acquisitions in
the lpfc driver, change load_flag into an unsigned long bitmask and use
clear_bit/test_bit bitwise atomic APIs instead of reliance on shost_lock
for synchronization.
Also, correct the test for FC_UNLOADING in lpfc_ct_handle_mibreq, which
incorrectly tests vport->fc_flag rather than vport->load_flag.
Justin Tee [Wed, 31 Jan 2024 18:51:09 +0000 (10:51 -0800)]
scsi: lpfc: Change lpfc_vport fc_flag member into a bitmask
In attempt to reduce the amount of unnecessary shost_lock acquisitions in
the lpfc driver, change fc_flag into an unsigned long bitmask and use
clear_bit/test_bit bitwise atomic APIs instead of reliance on shost_lock
for synchronization.
Justin Tee [Wed, 31 Jan 2024 18:51:08 +0000 (10:51 -0800)]
scsi: lpfc: Protect vport fc_nodes list with an explicit spin lock
In attempt to reduce the amount of unnecessary shost_lock acquisitions in
the lpfc driver, replace shost_lock with an explicit fc_nodes_list_lock
spinlock when accessing vport->fc_nodes lists. Although vport memory
region is owned by shost->hostdata, it is driver private memory and an
explicit fc_nodes list lock for fc_nodes list mutations is more appropriate
than locking the entire shost.
Justin Tee [Wed, 31 Jan 2024 18:51:07 +0000 (10:51 -0800)]
scsi: lpfc: Change nlp state statistic counters into atomic_t
There is no reason to use the shost_lock to synchronize an LLDD statistics
counter. Convert all the nlp state statistic counters into atomic_t.
Corresponding zeroing, increments, and reads are converted to atomic
versions.
Justin Tee [Wed, 31 Jan 2024 18:51:06 +0000 (10:51 -0800)]
scsi: lpfc: Remove shost_lock protection for fc_host_port shost APIs
Desiring to reduce the amount of unnecessary shost_lock acquisitions in the
lpfc driver, it has been determined that there is no need for shost_lock
protection when retrieving fc_host port information because it is only for
display to user via sysfs.
Justin Tee [Wed, 31 Jan 2024 18:51:05 +0000 (10:51 -0800)]
scsi: lpfc: Move handling of reset congestion statistics events
The ACQE notification event to reset congestion statistics should be moved
into the specific lpfc_sli4_async_sli_evt() routine instead of being
processed from the generic lpfc_sli4_async_event_proc() routine.
Justin Tee [Wed, 31 Jan 2024 18:51:04 +0000 (10:51 -0800)]
scsi: lpfc: Save FPIN frequency statistics upon receipt of peer cgn notifications
FPIN frequency is provided by the fabric in peer congestion notifications.
Currently, the frequency is only logged in a message, but it should also be
saved into the phba's cgn_fpin statistics member for proper application
functionality.
Justin Tee [Wed, 31 Jan 2024 18:51:03 +0000 (10:51 -0800)]
scsi: lpfc: Add condition to delete ndlp object after sending BLS_RJT to an ABTS
The "Nodelist not empty" log message and an accompanying delay may be
observed when deleting an NPIV port or unloading the lpfc driver. This can
occur due to receipt of an ABTS for which there is no corresponding login
context or ndlp allocated. In such cases, the driver allocates a new ndlp
object to send a BLS_RJT after which the ndlp object unintentionally
remains in the NLP_STE_UNUSED_NODE state forever.
Add a check to conditionally remove ndlp's initial reference count when
queuing a BLS response. If the initial reference is removed, then set
the NLP_DROPPED flag to notify other code paths.
Justin Tee [Wed, 31 Jan 2024 18:51:02 +0000 (10:51 -0800)]
scsi: lpfc: Fix failure to delete vports when discovery is in progress
Requests to delete an NPIV port may fail repeatedly if the initial request
is received during discovery.
If the FC_UNLOADING load_flag is set, then skip CT response processing for
the physical port. This allows discovery processing for other lpfc_vport
objects to reach their cmpl routines before deleting the vport.
Justin Tee [Wed, 31 Jan 2024 18:51:01 +0000 (10:51 -0800)]
scsi: lpfc: Remove NLP_RCV_PLOGI early return during RSCN processing for ndlps
Upon first RSCN receipt of a target server's remote port that is initially
acting as an initiator function, the driver marks the ndlp->nlp_type as an
initiator role. Then later, when processing an RSCN for a target function
role switch, that ndlp remote port is permanently stuck as an initiator
role and can never transition to be discovered as an updated target role
function.
Remove the NLP_RCV_PLOGI early return if statement clause so that the
NLP_NPR_2B_DISC flag gets set. This allows for role change detections.
Justin Tee [Wed, 31 Jan 2024 18:51:00 +0000 (10:51 -0800)]
scsi: lpfc: Allow lpfc_plogi_confirm_nport() logic to execute for Fabric nodes
Remove the early return NLP_FABRIC check in lpfc_plogi_confirm_nport()
because it is possible for switch domain controllers to change WWPN.
As a result, allow lpfc_plogi_confirm_nport() to detect that a new ndlp
should be initialized in such cases. The old ndlp object will be cleaned
up when dev_loss_tmo callbk executes.
Justin Tee [Wed, 31 Jan 2024 18:50:57 +0000 (10:50 -0800)]
scsi: lpfc: Fix possible memory leak in lpfc_rcv_padisc()
The call to lpfc_sli4_resume_rpi() in lpfc_rcv_padisc() may return an
unsuccessful status. In such cases, the elsiocb is not issued, the
completion is not called, and thus the elsiocb resource is leaked.
Check return value after calling lpfc_sli4_resume_rpi() and conditionally
release the elsiocb resource.
Justin Tee [Wed, 31 Jan 2024 18:50:56 +0000 (10:50 -0800)]
scsi: lpfc: Initialize status local variable in lpfc_sli4_repost_sgl_list()
A static code analyzer tool indicates that the local variable called status
in the lpfc_sli4_repost_sgl_list() routine could be used to print garbage
uninitialized values in the routine's log message.
Mike Christie [Tue, 23 Jan 2024 00:22:20 +0000 (18:22 -0600)]
scsi: core: Add kunit tests for scsi_check_passthrough()
Add some kunit tests for scsi_check_passthrough() so we can easily make
sure we are hitting the cases for which it's difficult to replicate in
hardware or even scsi_debug.
Mike Christie [Tue, 23 Jan 2024 00:22:18 +0000 (18:22 -0600)]
scsi: sr: Have midlayer retry get_sectorsize() errors
This has get_sectorsize() have the SCSI midlayer retry errors instead of
driving them itself.
There is one behavior change where we no longer retry when
scsi_execute_cmd() returns < 0, but we should be ok. We don't need to retry
for failures like the queue being removed, and for the case where there are
no tags/reqs the block layer waits/retries for us. For possible memory
allocation failures from blk_rq_map_kern() we use GFP_NOIO, so retrying
will probably not help.
Mike Christie [Tue, 23 Jan 2024 00:22:16 +0000 (18:22 -0600)]
scsi: sd: Have midlayer retry read_capacity_10() errors
This has read_capacity_10() have the SCSI midlayer retry errors instead of
driving them itself.
There are 2 behavior changes with this patch:
1. There is one behavior change where we no longer retry when
scsi_execute_cmd() returns < 0, but we should be ok. We don't need to
retry for failures like the queue being removed, and for the case where
there are no tags/reqs since the block layer waits/retries for us. For
possible memory allocation failures from blk_rq_map_kern() we use
GFP_NOIO, so retrying will probably not help.
2. For the specific UAs we checked for and retried, we would get
READ_CAPACITY_RETRIES_ON_RESET retries plus whatever retries were left
from the main loop's retries. Each UA now gets
READ_CAPACITY_RETRIES_ON_RESET retries, and the other errors get up to
3 retries. This is most likely ok, because
READ_CAPACITY_RETRIES_ON_RESET is already 10 and is not based on
anything specific like a spec or device, so the extra 3 we got from the
main loop was probably just an accident and is not going to help.
Mike Christie [Tue, 23 Jan 2024 00:22:15 +0000 (18:22 -0600)]
scsi: sd: Have pr commands retry UAs
It's common to get a UA when doing PR commands. It could be due to a target
restarting, transport level relogin or other PR commands like a release
causing it. The upper layers don't get the sense and in some cases have no
idea if it's a SCSI device, so this has the sd layer retry.
Mike Christie [Tue, 23 Jan 2024 00:22:14 +0000 (18:22 -0600)]
scsi: core: Have SCSI midlayer retry scsi_report_lun_scan() errors
This has scsi_report_lun_scan() have the SCSI midlayer retry errors instead
of driving them itself.
There is one behavior change where we no longer retry when
scsi_execute_cmd() returns < 0, but we should be ok. We don't need to retry
for failures like the queue being removed, and for the case where there are
no tags/reqs the block layer waits/retries for us. For possible memory
allocation failures from blk_rq_map_kern() we use GFP_NOIO, so retrying
will probably not help.
Mike Christie [Tue, 23 Jan 2024 00:22:11 +0000 (18:22 -0600)]
scsi: ch: Remove unit_attention
unit_attention is not used so remove it.
Signed-off-by: Mike Christie <michael.christie@oracle.com> Link: https://lore.kernel.org/r/20240123002220.129141-11-michael.christie@oracle.com Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: John Garry <john.g.garry@oracle.com> Reviewed-by: Martin Wilck <mwilck@suse.com> Acked-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Mike Christie [Tue, 23 Jan 2024 00:22:10 +0000 (18:22 -0600)]
scsi: sd: Have midlayer retry sd_sync_cache() errors
This has sd_sync_cache() have the SCSI midlayer retry errors instead of
driving them itself.
There is one behavior change where we no longer retry when
scsi_execute_cmd() returns < 0, but we should be ok. We don't need to retry
for failures like the queue being removed, and for the case where there are
no tags/reqs the block layer waits/retries for us. For possible memory
allocation failures from blk_rq_map_kern() we use GFP_NOIO, so retrying
will probably not help.
Mike Christie [Tue, 23 Jan 2024 00:22:06 +0000 (18:22 -0600)]
scsi: sd: Have midlayer retry sd_spinup_disk() errors
This simplifies sd_spinup_disk() so the SCSI midlayer retries errors for
it. Note that we retried every UA except Medium Not Present and also if
scsi_status_is_good() returned failed which would happen for all check
conditions. In this patch we use SCMD_FAILURE_STAT_ANY which will trigger
for the same conditions as when scsi_status_is_good() returns false and
there is status. This will cover all CCs including UAs so there is no
explicit failures array entry for UAs except for Medium Not Present which
we don't want to retry.
There is one behavior change where we no longer retry when
scsi_execute_cmd() returns < 0, but we should be ok. We don't need to retry
for failures like the queue being removed, and for the case where there are
no tags/reqs the block layer waits/retries for us. For possible memory
allocation failures from blk_rq_map_kern() we use GFP_NOIO, so retrying
will probably not help.
We do not handle the outside loop's retries because we want to sleep
between tries and we don't support that yet.
Mike Christie [Tue, 23 Jan 2024 00:22:05 +0000 (18:22 -0600)]
scsi: sd: Use separate buf for START_STOP in sd_spinup_disk()
We currently reuse the cmd buffer for the TUR and START_STOP commands
which requires us to reset the buffer when retrying. This has us use
separate buffers for the 2 commands so we can make them const and I think
it makes it easier to handle for retries but does not add too much extra to
the stack use.
Signed-off-by: Mike Christie <michael.christie@oracle.com> Link: https://lore.kernel.org/r/20240123002220.129141-5-michael.christie@oracle.com Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: John Garry <john.g.garry@oracle.com> Reviewed-by: Martin Wilck <mwilck@suse.com> Acked-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Mike Christie [Tue, 23 Jan 2024 00:22:04 +0000 (18:22 -0600)]
scsi: core: Retry INQUIRY after timeout
Description from: Martin Wilck <mwilck@suse.com>:
The SCSI mid layer doesn't retry commands after DID_TIME_OUT (see
scsi_noretry_cmd()). Packet loss in the fabric can cause spurious timeouts
during SCSI device probing, causing device probing to fail. This has been
observed in FCoE uplink failover tests, for example.
This patch fixes the issue by retrying the INQUIRY.
Signed-off-by: Mike Christie <michael.christie@oracle.com> Link: https://lore.kernel.org/r/20240123002220.129141-4-michael.christie@oracle.com Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin Wilck <mwilck@suse.com> Acked-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Mike Christie [Tue, 23 Jan 2024 00:22:02 +0000 (18:22 -0600)]
scsi: core: Allow passthrough to request midlayer retries
For passthrough we don't retry any error which we get a check condition
for. This results in a lot of callers driving their own retries for all
UAs, specific UAs, NOT_READY, specific sense values or any type of failure.
This adds the core code to allow passthrough users to specify what errors
they want the SCSI midlayer to retry for them. We can then convert users to
drop a lot of their sense parsing and retry handling.
Li Zhijian [Tue, 16 Jan 2024 04:51:43 +0000 (12:51 +0800)]
scsi: pm8001: Convert snprintf() to sysfs_emit()
Per filesystems/sysfs.rst, show() should only use sysfs_emit() or
sysfs_emit_at() when formatting the value to be returned to user space.
coccinelle complains that there are still a couple of functions that use
snprintf(). Convert them to sysfs_emit().
> ./drivers/scsi/pm8001/pm8001_ctl.c:883:8-16: WARNING: please use sysfs_emit
No functional change intended
CC: Jack Wang <jinpu.wang@cloud.ionos.com> CC: James E.J. Bottomley <jejb@linux.ibm.com> CC: Martin K. Petersen <martin.petersen@oracle.com> CC: linux-scsi@vger.kernel.org Signed-off-by: Li Zhijian <lizhijian@fujitsu.com> Link: https://lore.kernel.org/r/20240116045151.3940401-32-lizhijian@fujitsu.com Acked-by: Jack Wang <jinpu.wang@ionos.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Li Zhijian [Tue, 16 Jan 2024 04:51:36 +0000 (12:51 +0800)]
scsi: isci: Convert snprintf() to sysfs_emit()
Per filesystems/sysfs.rst, show() should only use sysfs_emit() or
sysfs_emit_at() when formatting the value to be returned to user space.
coccinelle complains that there are still a couple of functions that use
snprintf(). Convert them to sysfs_emit().
> ./drivers/scsi/isci/init.c:140:8-16: WARNING: please use sysfs_emit
No functional change intended
CC: Artur Paszkiewicz <artur.paszkiewicz@intel.com> CC: James E.J. Bottomley <jejb@linux.ibm.com> CC: Martin K. Petersen <martin.petersen@oracle.com> CC: linux-scsi@vger.kernel.org Signed-off-by: Li Zhijian <lizhijian@fujitsu.com> Link: https://lore.kernel.org/r/20240116045151.3940401-25-lizhijian@fujitsu.com Reviewed-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Li Zhijian [Tue, 16 Jan 2024 04:51:35 +0000 (12:51 +0800)]
scsi: ibmvscsi_tgt: Convert snprintf() to sysfs_emit()
Per filesystems/sysfs.rst, show() should only use sysfs_emit() or
sysfs_emit_at() when formatting the value to be returned to user space.
coccinelle complains that there are still a couple of functions that use
snprintf(). Convert them to sysfs_emit().
> ./drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c:3619:8-16: WARNING: please use sysfs_emit
> ./drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c:3625:8-16: WARNING: please use sysfs_emit
> ./drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c:3633:8-16: WARNING: please use sysfs_emit
No functional change intended
CC: Michael Cyr <mikecyr@linux.ibm.com> CC: James E.J. Bottomley <jejb@linux.ibm.com> CC: Martin K. Petersen <martin.petersen@oracle.com> CC: linux-scsi@vger.kernel.org CC: target-devel@vger.kernel.org Signed-off-by: Li Zhijian <lizhijian@fujitsu.com> Link: https://lore.kernel.org/r/20240116045151.3940401-24-lizhijian@fujitsu.com Acked-by: Tyrel Datwyler <tyreld@linux.ibm.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Lee Jones [Thu, 11 Jan 2024 13:17:26 +0000 (13:17 +0000)]
scsi: aacraid: aachba: Replace snprintf() with the safer scnprintf() variant
There is a general misunderstanding amongst engineers that {v}snprintf()
returns the length of the data *actually* encoded into the destination
array. However, as per the C99 standard {v}snprintf() really returns
the length of the data that *would have been* written if there were
enough space for it. This misunderstanding has led to buffer-overruns
in the past. It's generally considered safer to use the {v}scnprintf()
variants in their place (or even sprintf() in simple cases). So let's
do that.
Lee Jones [Thu, 11 Jan 2024 13:17:25 +0000 (13:17 +0000)]
scsi: 53c700: Remove snprintf() from sysfs call-backs and replace with sysfs_emit()
Since snprintf() has the documented, but still rather strange trait of
returning the length of the data that *would have been* written to the
array if space were available, rather than the arguably more useful
length of data *actually* written, it is usually considered wise to use
something else instead in order to avoid confusion.
In the case of sysfs call-backs, new wrappers exist that do just that.
Lee Jones [Thu, 11 Jan 2024 13:17:24 +0000 (13:17 +0000)]
scsi: 3w-xxxx: Remove snprintf() from sysfs call-backs and replace with sysfs_emit()
Since snprintf() has the documented, but still rather strange trait of
returning the length of the data that *would have been* written to the
array if space were available, rather than the arguably more useful
length of data *actually* written, it is usually considered wise to use
something else instead in order to avoid confusion.
In the case of sysfs call-backs, new wrappers exist that do just that.
Lee Jones [Thu, 11 Jan 2024 13:17:23 +0000 (13:17 +0000)]
scsi: 3w-sas: Remove snprintf() from sysfs call-backs and replace with sysfs_emit()
Since snprintf() has the documented, but still rather strange trait of
returning the length of the data that *would have been* written to the
array if space were available, rather than the arguably more useful
length of data *actually* written, it is usually considered wise to use
something else instead in order to avoid confusion.
In the case of sysfs call-backs, new wrappers exist that do just that.
Lee Jones [Thu, 11 Jan 2024 13:17:22 +0000 (13:17 +0000)]
scsi: 3w-9xxx: Remove snprintf() from sysfs call-backs and replace with sysfs_emit()
Since snprintf() has the documented, but still rather strange trait of
returning the length of the data that *would have been* written to the
array if space were available, rather than the arguably more useful
length of data *actually* written, it is usually considered wise to use
something else instead in order to avoid confusion.
In the case of sysfs call-backs, new wrappers exist that do just that.
Ranjan Kumar [Thu, 28 Dec 2023 11:48:09 +0000 (17:18 +0530)]
scsi: mpt3sas: Reload SBR without rebooting HBA
Add a new IOCTL command MPT3ENABLEDIAGSBRRELOAD. As a part of firmware
update operation, applications use this IOCTL command to set the SBR reload
bit in the Host Diagnostic register. This permits HBA firmware to be
updated without powercycling the system.
Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202312280909.MZyhxwBL-lkp@intel.com/ Closes: https://lore.kernel.org/oe-kbuild-all/202312281141.jDyPezRn-lkp@intel.com/ Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com> Link: https://lore.kernel.org/r/20231228114810.11923-2-ranjan.kumar@broadcom.com Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Eric Chanudet [Tue, 23 Jan 2024 19:28:57 +0000 (14:28 -0500)]
scsi: ufs: qcom: Avoid re-init quirk when gears match
On sa8775p-ride, probing the HBA will go through the
UFSHCD_QUIRK_REINIT_AFTER_MAX_GEAR_SWITCH path although the power info is
the same during the second init.
The REINIT quirk only applies starting with controller v4. For these,
ufs_qcom_get_hs_gear() reads the highest supported gear when setting the
host_params. After the negotiation, if the host and device are on the same
gear, it is the highest gear supported between the two. Skip REINIT to save
some time.
Signed-off-by: Eric Chanudet <echanude@redhat.com> Link: https://lore.kernel.org/r/20240123192854.1724905-4-echanude@redhat.com Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Tested-by: Andrew Halaney <ahalaney@redhat.com> # sa8775p-ride Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Andrew Halaney [Tue, 23 Jan 2024 19:13:36 +0000 (13:13 -0600)]
scsi: ufs: qcom: Clarify comments about the initial phy_gear
The comments that currently are within the hw_ver < 4 conditional are
misleading. They really apply to various branches of the conditionals there
and incorrectly state that the phy_gear value can increase.
Right now the logic is to:
- Default to max supported gear for phy_gear
- Set phy_gear to minimum value if version < 4 since those versions only
support one PHY init sequence (and therefore don't need reinit)
- Set phy_gear to the optimal value if the device version is already
populated in the controller registers on boot
Let's move some of the comment to outside the if statement and clean up the
bit left about switching to a higher gear on reinit. This way the comment
more accurately reflects the logic.
Xiang Chen [Mon, 22 Jan 2024 06:25:47 +0000 (14:25 +0800)]
scsi: hisi_sas: Remove hisi_hba->timer for v3 hw
hisi_hba->timer is not used for v3 hw but there are two places that some
operations related to hisi_hba->timer are called by v3 hw:
- Deleting the timer in function hisi_sas_v3_hw() which is only for v3 hw;
- Deleting the timer in function hisi_sas_controller_reset_prepare() which
is common for v1/v2/v3 hw.
We can remove the timer in the first case, but for the second scenario we
need to remove it only for v3 hw, so check hw->sht which is NULL only for
v3 hw before deleting hisi_hba->timer.
Yihang Li [Mon, 22 Jan 2024 06:25:45 +0000 (14:25 +0800)]
scsi: hisi_sas: Remove redundant checks for automatic debugfs dump
In commit 63f0733d07ce ("scsi: hisi_sas: Allocate DFX memory during dump
trigger"), the memory allocation time of the DFX is changed from device
initialization to dump occurs, so .debugfs_itct is not a valid address and
do not need to check.
The parameter hisi_sas_debugfs_enable is enough to check whether automatic
debugfs dump is triggered, so remove redunant checks.
Yihang Li [Mon, 22 Jan 2024 06:25:44 +0000 (14:25 +0800)]
scsi: hisi_sas: Fix a deadlock issue related to automatic dump
If we issue a disabling PHY command, the device attached with it will go
offline, if a 2 bit ECC error occurs at the same time, a hung task may be
found:
This is because when the device goes offline, we obtain the hisi_hba
semaphore and send the ABORT_DEV command to the device. However, the
internal abort timed out due to the 2 bit ECC error and triggers automatic
dump. In addition, since the hisi_hba semaphore has been obtained, the dump
cannot be executed and the controller cannot be reset.
Therefore, the deadlocks occur on the following circular dependencies:
hisi_sas_dev_gone() -> down() -> hisi_sas_internal_task_abort_dev() -> ...
-> hisi_sas_internal_abort_timeout() -> down().
The deadlock is triggered only when the timeout occurs during device goes
offline. To fix this issue, use .rst_ha_timeout to distinguish the scenario
where a device goes offline from other scenarios.
Fixes: 2ff07b5c6fe9 ("scsi: hisi_sas: Directly call register snapshot instead of using workqueue") Signed-off-by: Yihang Li <liyihang9@huawei.com> Signed-off-by: Xiang Chen <chenxiang66@hisilicon.com> Link: https://lore.kernel.org/r/1705904747-62186-2-git-send-email-chenxiang66@hisilicon.com Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Justin Stitt [Tue, 12 Dec 2023 01:20:20 +0000 (01:20 +0000)]
scsi: ibmvscsi_tgt: Replace deprecated strncpy() with strscpy()
strncpy() is deprecated for use on NUL-terminated destination strings [1]
and as such we should prefer more robust and less ambiguous string
interfaces.
We don't need the NUL-padding behavior that strncpy() provides as vscsi is
NUL-allocated in ibmvscsis_probe() which proceeds to call
ibmvscsis_adapter_info():
Following the same idea, `partition_name` is defiend as:
| static char partition_name[PARTITION_NAMELEN] = "UNKNOWN";
... which is NUL-padded already, meaning strscpy() is the best option.
Considering the above, a suitable replacement is strscpy() [2] due to the
fact that it guarantees NUL-termination on the destination buffer without
unnecessarily NUL-padding.
However, for cap->name and info let's use strscpy_pad() as they are
allocated via dma_alloc_coherent():
| cap = dma_alloc_coherent(&vscsi->dma_dev->dev, olen, &token,
| GFP_ATOMIC);
&
| info = dma_alloc_coherent(&vscsi->dma_dev->dev, sizeof(*info), &token,
| GFP_ATOMIC);
ChanWoo Lee [Tue, 2 Jan 2024 01:42:22 +0000 (10:42 +0900)]
scsi: ufs: mcq: Add definition for REG_UFS_MEM_CFG register
Instead of hardcoding the register field, add the proper definition. While
at it, let's also use ufshcd_rmwl() to simplify updating this register.
Reviewed-by: Peter Wang <peter.wang@mediatek.com> Signed-off-by: ChanWoo Lee <cw9316.lee@samsung.com> Link: https://lore.kernel.org/r/20240102014222.23351-1-cw9316.lee@samsung.com Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Petr Mladek [Thu, 11 Jan 2024 16:24:19 +0000 (17:24 +0100)]
scsi: core: Safe warning about bad dev info string
Both "model" and "strflags" are passed to "%s" even when one or both are
NULL.
It is safe because vsprintf() would detect the NULL pointer and print
"(null)". But it is a kernel-specific feature and compiler warns about it:
<warning>
In file included from include/linux/kernel.h:19,
from arch/x86/include/asm/percpu.h:27,
from arch/x86/include/asm/current.h:6,
from include/linux/sched.h:12,
from include/linux/blkdev.h:5,
from drivers/scsi/scsi_devinfo.c:3:
drivers/scsi/scsi_devinfo.c: In function 'scsi_dev_info_list_add_str':
>> include/linux/printk.h:434:44: warning: '%s' directive argument is null [-Wformat-overflow=]
434 | #define printk(fmt, ...) printk_index_wrap(_printk, fmt, ##__VA_ARGS__)
| ^
include/linux/printk.h:430:3: note: in definition of macro 'printk_index_wrap'
430 | _p_func(_fmt, ##__VA_ARGS__); \
| ^~~~~~~
drivers/scsi/scsi_devinfo.c:551:4: note: in expansion of macro 'printk'
551 | printk(KERN_ERR "%s: bad dev info string '%s' '%s'"
| ^~~~~~
drivers/scsi/scsi_devinfo.c:552:14: note: format string is defined here
552 | " '%s'\n", __func__, vendor, model,
| ^~
</warning>
Do not rely on the kernel specific behavior and print the message a safe
way.
Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202401112002.AOjwMNM0-lkp@intel.com/ Signed-off-by: Petr Mladek <pmladek@suse.com> Link: https://lore.kernel.org/r/20240111162419.12406-1-pmladek@suse.com Reviewed-by: Bart Van Assche <bvanassche@acm.org> Acked-by: Chris Down <chris@chrisdown.name> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Peter Wang [Tue, 9 Jan 2024 12:40:14 +0000 (20:40 +0800)]
scsi: core: Move autosuspend timer delay to Scsi_Host
The runtime suspend timer delay is a const value in scsi_host_template
which a host driver cannot modify at runtime. Move the delay to Scsi_Host
to allow a driver to update it.
Maramaina Naresh [Tue, 19 Dec 2023 12:37:06 +0000 (18:07 +0530)]
scsi: ufs: ufs-mediatek: Migrate to UFSHCD generic CPU latency PM QoS support
The PM QoS feature found in the MediaTek UFS driver was moved to the UFSHCD
core. Hence remove it from MediaTek UFS driver as it is redundant now.
Reviewed-by: Peter Wang <peter.wang@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Signed-off-by: Maramaina Naresh <quic_mnaresh@quicinc.com> Link: https://lore.kernel.org/r/20231219123706.6463-3-quic_mnaresh@quicinc.com Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Maramaina Naresh [Tue, 19 Dec 2023 12:37:05 +0000 (18:07 +0530)]
scsi: ufs: core: Add CPU latency QoS support for UFS driver
Register UFS driver to CPU latency PM QoS framework to improve UFS device
random I/O performance.
PM QoS initialization will insert new QoS request into the CPU latency QoS
list with the maximum latency PM_QOS_DEFAULT_VALUE value.
The UFS driver will vote for performance mode on scale up and power save
mode for scale down.
If clock scaling feature is not enabled then voting will be based on clock
on or off condition. Also provide a sysfs interface to enable/disable PM
QoS feature.
tiotest benchmark tool I/O performance results on sm8550 platform:
1. Without PM QoS support
Type (Speed in) | Average of 18 iterations
Random Write(IPOS) | 41065.13
Random Read(IPOS) | 37101.3
2. With PM QoS support
Type (Speed in) | Average of 18 iterations
Random Write(IPOS) | 46784.9
Random Read(IPOS) | 42943.4
(Improvement with PM QoS = ~15%).
Reviewed-by: Peter Wang <peter.wang@mediatek.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Co-developed-by: Nitin Rawat <quic_nitirawa@quicinc.com> Signed-off-by: Nitin Rawat <quic_nitirawa@quicinc.com> Co-developed-by: Naveen Kumar Goud Arepalli <quic_narepall@quicinc.com> Signed-off-by: Naveen Kumar Goud Arepalli <quic_narepall@quicinc.com> Signed-off-by: Maramaina Naresh <quic_mnaresh@quicinc.com> Link: https://lore.kernel.org/r/20231219123706.6463-2-quic_mnaresh@quicinc.com Reviewed-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Linus Torvalds [Sun, 21 Jan 2024 22:01:12 +0000 (14:01 -0800)]
Merge tag 'bcachefs-2024-01-21' of https://evilpiepirate.org/git/bcachefs
Pull more bcachefs updates from Kent Overstreet:
"Some fixes, Some refactoring, some minor features:
- Assorted prep work for disk space accounting rewrite
- BTREE_TRIGGER_ATOMIC: after combining our trigger callbacks, this
makes our trigger context more explicit
- A few fixes to avoid excessive transaction restarts on
multithreaded workloads: fstests (in addition to ktest tests) are
now checking slowpath counters, and that's shaking out a few bugs
- Assorted tracepoint improvements
- Starting to break up bcachefs_format.h and move on disk types so
they're with the code they belong to; this will make room to start
documenting the on disk format better.
- A few minor fixes"
* tag 'bcachefs-2024-01-21' of https://evilpiepirate.org/git/bcachefs: (46 commits)
bcachefs: Improve inode_to_text()
bcachefs: logged_ops_format.h
bcachefs: reflink_format.h
bcachefs; extents_format.h
bcachefs: ec_format.h
bcachefs: subvolume_format.h
bcachefs: snapshot_format.h
bcachefs: alloc_background_format.h
bcachefs: xattr_format.h
bcachefs: dirent_format.h
bcachefs: inode_format.h
bcachefs; quota_format.h
bcachefs: sb-counters_format.h
bcachefs: counters.c -> sb-counters.c
bcachefs: comment bch_subvolume
bcachefs: bch_snapshot::btime
bcachefs: add missing __GFP_NOWARN
bcachefs: opts->compression can now also be applied in the background
bcachefs: Prep work for variable size btree node buffers
bcachefs: grab s_umount only if snapshotting
...
Linus Torvalds [Sun, 21 Jan 2024 19:14:40 +0000 (11:14 -0800)]
Merge tag 'timers-core-2024-01-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer updates from Thomas Gleixner:
"Updates for time and clocksources:
- A fix for the idle and iowait time accounting vs CPU hotplug.
The time is reset on CPU hotplug which makes the accumulated
systemwide time jump backwards.
- Assorted fixes and improvements for clocksource/event drivers"
* tag 'timers-core-2024-01-21' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
tick-sched: Fix idle and iowait sleeptime accounting vs CPU hotplug
clocksource/drivers/ep93xx: Fix error handling during probe
clocksource/drivers/cadence-ttc: Fix some kernel-doc warnings
clocksource/drivers/timer-ti-dm: Fix make W=n kerneldoc warnings
clocksource/timer-riscv: Add riscv_clock_shutdown callback
dt-bindings: timer: Add StarFive JH8100 clint
dt-bindings: timer: thead,c900-aclint-mtimer: separate mtime and mtimecmp regs
projects / users / willy / xarray.git / log