libata-eh cleanups
Subsystems (2)
Damien Le Moal
maintainerv2 cover libata-eh cleanups
3 patches to cleanup libata-eh code and its documentation. No functional changes are introduced.
Changes from v1:
- Correct compilation error when CONFIG_ATA_SFF is not enabled by defining a stub for sata_sff_hardreset().
Damien Le Moal (3):
ata: libata-eh: Make ata_eh_followup_srst_needed() return a bool
ata: libata-eh: Remove ata_do_eh()
Documentation: driver-api: Update libata error handler information
Documentation/driver-api/libata.rst | 21 ++++++----- drivers/ata/libata-eh.c | 56 ++++++++++------------------- drivers/ata/libata-sff.c | 10 +----- include/linux/libata.h | 11 ++++-- 4 files changed, 39 insertions(+), 59 deletions(-)
Signature · 1 line
2.50.0
Damien Le Moal
maintainerPATCH v2 1/3 ata: libata-eh: Make ata_eh_followup_srst_needed() return a bool
ata_eh_followup_srst_needed() returns an integer used as a boolean. So change this function to return that type.
No functional changes.
--- drivers/ata/libata-eh.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/ata/libata-eh.c +5 -5
@@ -2781,15 +2781,15 @@ static int ata_do_reset(struct ata_link *link, ata_reset_fn_t reset,return reset (link, classes, deadline); } -static int ata_eh_followup_srst_needed (struct ata_link *link,int rc) +static bool ata_eh_followup_srst_needed (struct ata_link *link,int rc) {if ((link->flags &ATA_LFLAG_NO_SRST ) ||ata_link_offline (link)) -return 0 ; +return false;if (rc == -EAGAIN ) -return 1 ; +return true;if (sata_pmp_supported (link->ap) &&ata_is_host_link (link)) -return 1 ; -return 0 ; +return true; +return false; } int ata_eh_reset(struct ata_link *link, int classify,
Signature · 1 line
2.50.0
Damien Le Moal
maintainerPATCH v2 2/3 ata: libata-eh: Remove ata_do_eh()
In reply to Damien Le MoalThe only reason for ata_do_eh() to exist is that the two caller sites, ata_std_error_handler() and ata_sff_error_handler() may pass to it a NULL hardreset operation so that the built-in (generic) hardreset operation for a driver is ignored if the adapter SCR access is not available.
However, ata_std_error_handler() and ata_sff_error_handler() modifications of the hardreset port operation can easily be combined as they are mutually exclusive. That is, a driver using sata_std_hardreset() as its hardreset operation cannot use sata_sff_hardreset() and vice-versa.
With this observation, ata_do_eh() can be removed and its code moved to ata_std_error_handler(). The condition used to ignore the builtin hardreset port operations is modified to be the one that was used in ata_sff_error_handler(). This requires defining a stub for the function sata_sff_hardreset() to avoid compilation errors when CONFIG_ATA_SFF is not enabled.
This change simplifies ata_sff_error_handler() as this function now only needs to call ata_std_error_handler().
No functional changes.
--- drivers/ata/libata-eh.c | 46 ++++++++++++---------------------------- drivers/ata/libata-sff.c | 10 +-------- include/linux/libata.h | 11 +++++++--- 3 files changed, 22 insertions(+), 45 deletions(-)
drivers/ata/libata-eh.c +13 -33
@@ -4067,59 +4067,39 @@ void ata_eh_finish(struct ata_port *ap) }/** -* ata_do_eh - do standard error handling +* ata_std_error_handler - standard error handler * @ap: host port to handle error for * -* @prereset: prereset method (can be NULL) -* @softreset: softreset method (can be NULL) -* @hardreset: hardreset method (can be NULL) -* @postreset: postreset method (can be NULL) -* * Perform standard error handling sequence. * * LOCKING: * Kernel thread context (may sleep). */ -void ata_do_eh (struct ata_port *ap,ata_prereset_fn_t prereset, -ata_reset_fn_t softreset,ata_reset_fn_t hardreset, -ata_postreset_fn_t postreset) +void ata_std_error_handler (struct ata_port *ap) { -struct ata_device *dev; +struct ata_port_operations *ops = ap->ops; +ata_reset_fn_t hardreset = ops->hardreset;int rc; +/* Ignore built-in hardresets if SCR access is not available */ +if ((hardreset == sata_std_hardreset || + hardreset == sata_sff_hardreset) && !sata_scr_valid (&ap->link)) + hardreset =NULL ; +ata_eh_autopsy (ap);ata_eh_report (ap); - rc =ata_eh_recover (ap, prereset, softreset, hardreset, postreset, -NULL ); + rc =ata_eh_recover (ap, ops->prereset, ops->softreset, + hardreset, ops->postreset,NULL );if (rc) { +struct ata_device *dev; +ata_for_each_dev (dev, &ap->link,ALL )ata_dev_disable (dev); }ata_eh_finish (ap); } - -/** -* ata_std_error_handler - standard error handler -* @ap: host port to handle error for -* -* Standard error handler -* -* LOCKING: -* Kernel thread context (may sleep). -*/ -void ata_std_error_handler (struct ata_port *ap) -{ -struct ata_port_operations *ops = ap->ops; -ata_reset_fn_t hardreset = ops->hardreset; - -/* ignore built-in hardreset if SCR access is not available */ -if (hardreset == sata_std_hardreset && !sata_scr_valid (&ap->link)) - hardreset =NULL ; - -ata_do_eh (ap, ops->prereset, ops->softreset, hardreset, ops->postreset); -}EXPORT_SYMBOL_GPL (ata_std_error_handler);#ifdef CONFIG_PM
drivers/ata/libata-sff.c +1 -9
@@ -2054,8 +2054,6 @@ EXPORT_SYMBOL_GPL(ata_sff_drain_fifo);*/ void ata_sff_error_handler (struct ata_port *ap) { -ata_reset_fn_t softreset = ap->ops->softreset; -ata_reset_fn_t hardreset = ap->ops->hardreset;struct ata_queued_cmd *qc;unsigned long flags;
@@ -2077,13 +2075,7 @@ void ata_sff_error_handler(struct ata_port *ap)spin_unlock_irqrestore (ap->lock, flags); -/* ignore built-in hardresets if SCR access is not available */ -if ((hardreset == sata_std_hardreset || - hardreset == sata_sff_hardreset) && !sata_scr_valid (&ap->link)) - hardreset =NULL ; - -ata_do_eh (ap, ap->ops->prereset, softreset, hardreset, - ap->ops->postreset); +ata_std_error_handler (ap); }EXPORT_SYMBOL_GPL (ata_sff_error_handler);
include/linux/libata.h +8 -3
@@ -1412,9 +1412,6 @@ extern void ata_eh_thaw_port(struct ata_port *ap);extern void ata_eh_qc_complete (struct ata_queued_cmd *qc);extern void ata_eh_qc_retry (struct ata_queued_cmd *qc); -extern void ata_do_eh (struct ata_port *ap,ata_prereset_fn_t prereset, -ata_reset_fn_t softreset,ata_reset_fn_t hardreset, -ata_postreset_fn_t postreset);extern void ata_std_error_handler (struct ata_port *ap);extern void ata_std_sched_eh (struct ata_port *ap);extern void ata_std_end_eh (struct ata_port *ap);
@@ -2152,6 +2149,14 @@ static inline u8 ata_wait_idle(struct ata_port *ap)return status; } +#else /* CONFIG_ATA_SFF */ + +static inline int sata_sff_hardreset (struct ata_link *link,unsigned int *class, +unsigned long deadline) +{ +return -EOPNOTSUPP ; +} +#endif /* CONFIG_ATA_SFF */ #endif /* __LINUX_LIBATA_H__ */
Signature · 1 line
2.50.0
Damien Le Moal
maintainerPATCH v2 3/3 Documentation: driver-api: Update libata error handler information
In reply to Damien Le MoalUpdate ``->error_handler()`` section of the libata documentation file Documentation/driver-api/libata.rst to remove the reference to the function ata_do_eh() as that function was removed. The reference to the function ata_bmdma_drive_eh() is also removed as that function does not exist at all. And while at it, cleanup the description of the various port reset operations using a bullet list.
--- Documentation/driver-api/libata.rst | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-)
Documentation/driver-api/libata.rst +12 -9
@@ -283,18 +283,21 @@ interrupts, start DMA engine, etc. ``->error_handler()`` is a driver's hook into probe, hotplug, and recovery and other exceptional conditions. The primary responsibility of an -implementation is to call :c:func:`ata_do_eh` or :c:func:`ata_bmdma_drive_eh` -with a set of EH hooks as arguments: +implementation is to call :c:func:`ata_std_error_handler`. -'prereset' hook (may be NULL) is called during an EH reset, before any -other actions are taken. +:c:func:`ata_std_error_handler` will call the various port reset operations as +needed. -'postreset' hook (may be NULL) is called after the EH reset is -performed. Based on existing conditions, severity of the problem, and -hardware capabilities, +* The 'prereset' operation (may be NULL) is called during an EH reset, before + any other action is taken. -Either 'softreset' (may be NULL) or 'hardreset' (may be NULL) will be -called to perform the low-level EH reset. +* The 'postreset' hook (may be NULL) is called after the EH reset is performed. + Based on existing conditions, severity of the problem, and hardware + capabilities, + +* Either the 'softreset' operation (may be NULL) or the 'hardreset' operation + (may be NULL) will be called to perform the low-level EH reset. If both + operations are defined, 'hardreset' is preferred and used. ::
Signature · 1 line
2.50.0
Re: [PATCH v2 1/3] ata: libata-eh: Make ata_eh_followup_srst_needed() return a bool
In reply to Damien Le MoalQuoted 25 lines · Damien Le Moal
ata_eh_followup_srst_needed()returns an integer used as a boolean. So change this function to return that type.No functional changes.
Signed-off-by Damien Le Moal <[email protected]>--- drivers/ata/libata-eh.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-)drivers/ata/libata-eh.c +2
-2@@ -2781,15 +2781,15 @@ static int ata_do_reset(struct ata_link *link, ata_reset_fn_t reset,return reset (link, classes, deadline); } -static int ata_eh_followup_srst_needed (struct ata_link *link,int rc) +static bool ata_eh_followup_srst_needed (struct ata_link *link,int rc) {if ((link->flags &ATA_LFLAG_NO_SRST ) ||ata_link_offline (link)) -return 0 ; +return false;if (rc == -EAGAIN )
drivers/ata/libata-eh.c static int ata_do_reset(struct ata_link *link, ata_reset_fn_t reset, · line 2795
Quoted 10 lines
drivers/ata/libata-eh.c +3
-3-return 1 ; +return true;if (sata_pmp_supported (link->ap) &&ata_is_host_link (link)) -return 1 ; -return 0 ; +return true; +return false; } int ata_eh_reset(struct ata_link *link, int classify,
Cheers,
Hannes
Re: [PATCH v2 2/3] ata: libata-eh: Remove ata_do_eh()
In reply to Damien Le MoalQuoted 22 lines · Damien Le Moal
The only reason for
ata_do_eh()to exist is that the two caller sites,ata_std_error_handler()andata_sff_error_handler()may pass to it a NULL hardreset operation so that the built-in (generic) hardreset operation for a driver is ignored if the adapter SCR access is not available.However,
ata_std_error_handler()andata_sff_error_handler()modifications of the hardreset port operation can easily be combined as they are mutually exclusive. That is, a driver usingsata_std_hardreset()as its hardreset operation cannot usesata_sff_hardreset()and vice-versa.With this observation,
ata_do_eh()can be removed and its code moved toata_std_error_handler(). The condition used to ignore the builtin hardreset port operations is modified to be the one that was used inata_sff_error_handler(). This requires defining a stub for the functionsata_sff_hardreset()to avoid compilation errors whenCONFIG_ATA_SFFis not enabled.This change simplifies
ata_sff_error_handler()as this function now only needs to callata_std_error_handler().
Quoted 9 lines
No functional changes.
Signed-off-by Damien Le Moal <[email protected]>--- drivers/ata/libata-eh.c | 46 ++++++++++++---------------------------- drivers/ata/libata-sff.c | 10 +-------- include/linux/libata.h | 11 +++++++--- 3 files changed, 22 insertions(+), 45 deletions(-)
Cheers,
Hannes
Re: [PATCH v2 3/3] Documentation: driver-api: Update libata error handler information
In reply to Damien Le MoalQuoted 7 lines · Damien Le Moal
Update ``->error_handler()`` section of the libata documentation file
Documentation/driver-api/libata.rstto remove the reference to the functionata_do_eh()as that function was removed. The reference to the functionata_bmdma_drive_eh()is also removed as that function does not exist at all. And while at it, cleanup the description of the various port reset operations using a bullet list.
Quoted 5 lines
Signed-off-by Damien Le Moal <[email protected]>--- Documentation/driver-api/libata.rst | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-)
Cheers,
Hannes
Re: [PATCH v2 3/3] Documentation: driver-api: Update libata error handler information
In reply to Damien Le MoalQuoted 34 lines · Damien Le Moal
Update ``->error_handler()`` section of the libata documentation file
Documentation/driver-api/libata.rstto remove the reference to the functionata_do_eh()as that function was removed. The reference to the functionata_bmdma_drive_eh()is also removed as that function does not exist at all. And while at it, cleanup the description of the various port reset operations using a bullet list.Signed-off-by Damien Le Moal <[email protected]>--- Documentation/driver-api/libata.rst | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-)Documentation/driver-api/libata.rst +5
-7@@ -283,18 +283,21 @@ interrupts, start DMA engine, etc. ``->error_handler()`` is a driver's hook into probe, hotplug, and recovery and other exceptional conditions. The primary responsibility of an -implementation is to call :c:func:`ata_do_eh` or :c:func:`ata_bmdma_drive_eh` -with a set of EH hooks as arguments: +implementation is to call :c:func:`ata_std_error_handler`. -'prereset' hook (may be NULL) is called during an EH reset, before any -other actions are taken. +:c:func:`ata_std_error_handler` will call the various port reset operations as +needed. -'postreset' hook (may be NULL) is called after the EH reset is -performed. Based on existing conditions, severity of the problem, and -hardware capabilities, +* The 'prereset' operation (may be NULL) is called during an EH reset, before + any other action is taken.
Documentation/driver-api/libata.rst interrupts, start DMA engine, etc. · line 300
Quoted 10 lines
Documentation/driver-api/libata.rst +7
-2-Either 'softreset' (may be NULL) or 'hardreset' (may be NULL) will be -called to perform the low-level EH reset. +* The 'postreset' hook (may be NULL) is called after the EH reset is performed. + Based on existing conditions, severity of the problem, and hardware + capabilities, + +* Either the 'softreset' operation (may be NULL) or the 'hardreset' operation + (may be NULL) will be called to perform the low-level EH reset. If both + operations are defined, 'hardreset' is preferred and used.
Since you are mentioning what happens when both are defined, should we mention what happens when none are defined? (None of the callbacks will be called.)
Re: [PATCH v2 1/3] ata: libata-eh: Make ata_eh_followup_srst_needed() return a bool
In reply to Damien Le MoalQuoted 6 lines · Damien Le Moal
ata_eh_followup_srst_needed()returns an integer used as a boolean. So change this function to return that type.No functional changes.
Signed-off-by Damien Le Moal <[email protected]>
Niklas Cassel
maintainerRe: [PATCH v2 2/3] ata: libata-eh: Remove ata_do_eh()
In reply to Damien Le MoalQuoted 2 lines · Damien Le Moal
The only reason for
ata_do_eh()to exist is that the two caller sites,ata_std_error_handler()andata_sff_error_handler()may pass to it a
s/may pass to it a/may pass it a/
Quoted 7 lines
NULL hardreset operation so that the built-in (generic) hardreset operation for a driver is ignored if the adapter SCR access is not available.
However,
ata_std_error_handler()andata_sff_error_handler()modifications of the hardreset port operation can easily be combined as they are mutually exclusive. That is, a driver using
sata_std_hardreset() can be on previous line without exceeding 75 columns.
Quoted 5 lines
sata_std_hardreset()as its hardreset operation cannot usesata_sff_hardreset()and vice-versa.With this observation,
ata_do_eh()can be removed and its code moved toata_std_error_handler(). The condition used to ignore the builtin
s/builtin/built-in/
Quoted 1 line
hardreset port operations is modified to be the one that was used in
s/operations/operations/
(you are ignoring the built-in hardreset operation, which is either sata_std_hardreset() or sata_sff_hardreset(), but still the function pointer will point to a single operation, so I think singular is more correct here.)
Quoted 42 lines
ata_sff_error_handler(). This requires defining a stub for the functionsata_sff_hardreset()to avoid compilation errors whenCONFIG_ATA_SFFis not enabled.This change simplifies
ata_sff_error_handler()as this function now only needs to callata_std_error_handler().No functional changes.
Signed-off-by Damien Le Moal <[email protected]>--- drivers/ata/libata-eh.c | 46 ++++++++++++---------------------------- drivers/ata/libata-sff.c | 10 +-------- include/linux/libata.h | 11 +++++++--- 3 files changed, 22 insertions(+), 45 deletions(-)drivers/ata/libata-eh.c +2
-9@@ -4067,59 +4067,39 @@ void ata_eh_finish(struct ata_port *ap) }/** -* ata_do_eh - do standard error handling +* ata_std_error_handler - standard error handler * @ap: host port to handle error for * -* @prereset: prereset method (can be NULL) -* @softreset: softreset method (can be NULL) -* @hardreset: hardreset method (can be NULL) -* @postreset: postreset method (can be NULL) -* * Perform standard error handling sequence. * * LOCKING: * Kernel thread context (may sleep). */ -void ata_do_eh (struct ata_port *ap,ata_prereset_fn_t prereset, -ata_reset_fn_t softreset,ata_reset_fn_t hardreset, -ata_postreset_fn_t postreset) +void ata_std_error_handler (struct ata_port *ap)
drivers/ata/libata-eh.c void ata_eh_finish(struct ata_port *ap) · line 4087
Quoted 10 lines
drivers/ata/libata-eh.c +6
-1{ -struct ata_device *dev; +struct ata_port_operations *ops = ap->ops; +ata_reset_fn_t hardreset = ops->hardreset;int rc; +/* Ignore built-in hardresets if SCR access is not available */ +if ((hardreset == sata_std_hardreset || + hardreset == sata_sff_hardreset) && !sata_scr_valid (&ap->link)) + hardreset =NULL ;
I think it would be nicer to just do:
if ((hardreset == sata_std_hardreset ||
hardreset == sata_sff_hardreset) && !sata_scr_valid(&ap->link))
link->flags |= ATA_LFLAG_NO_HRST;since ata_eh_reset() will already do
hardreset = NULL; if that flag is set.
This way, we can also simplify the function signature of ata_eh_recover() to:
int ata_eh_recover(struct ata_port *ap, bool use_pmp_ops,
struct ata_link **r_failed_link)
and ata_eh_reset() to:
int ata_eh_reset(struct ata_link *link, int classify, bool use_pmp_ops)
And then in:
int ata_eh_reset(struct ata_link *link, int classify, bool use_pmp_ops)
{
...
if (use_pmp_ops) {
prereset = ops->prereset;
softreset = ops->softreset;
hardreset = ops->hardreset;
postreset = ops->postreset;
} else {
prereset = ops->pmp_prereset;
softreset = ops->pmp_softreset;
hardreset = ops->pmp_hardreset;
postreset = ops->pmp_postreset;
}
if (link->flags & ATA_LFLAG_NO_HRST)
hardreset = NULL;
...
}(The if (link->flags & ATA_LFLAG_NO_HRST) statement is already there in ata_eh_reset().)
Quoted 84 lines
drivers/ata/libata-eh.c +5
-23+ata_eh_autopsy (ap);ata_eh_report (ap); - rc =ata_eh_recover (ap, prereset, softreset, hardreset, postreset, -NULL ); + rc =ata_eh_recover (ap, ops->prereset, ops->softreset, + hardreset, ops->postreset,NULL );if (rc) { +struct ata_device *dev; +ata_for_each_dev (dev, &ap->link,ALL )ata_dev_disable (dev); }ata_eh_finish (ap); } - -/** -* ata_std_error_handler - standard error handler -* @ap: host port to handle error for -* -* Standard error handler -* -* LOCKING: -* Kernel thread context (may sleep). -*/ -void ata_std_error_handler (struct ata_port *ap) -{ -struct ata_port_operations *ops = ap->ops; -ata_reset_fn_t hardreset = ops->hardreset; - -/* ignore built-in hardreset if SCR access is not available */ -if (hardreset == sata_std_hardreset && !sata_scr_valid (&ap->link)) - hardreset =NULL ; - -ata_do_eh (ap, ops->prereset, ops->softreset, hardreset, ops->postreset); -}EXPORT_SYMBOL_GPL (ata_std_error_handler);#ifdef CONFIG_PM drivers/ata/libata-sff.c +1
-9@@ -2054,8 +2054,6 @@ EXPORT_SYMBOL_GPL(ata_sff_drain_fifo);*/ void ata_sff_error_handler (struct ata_port *ap) { -ata_reset_fn_t softreset = ap->ops->softreset; -ata_reset_fn_t hardreset = ap->ops->hardreset;struct ata_queued_cmd *qc;unsigned long flags;@@ -2077,13 +2075,7 @@ void ata_sff_error_handler(struct ata_port *ap)spin_unlock_irqrestore (ap->lock, flags); -/* ignore built-in hardresets if SCR access is not available */ -if ((hardreset == sata_std_hardreset || - hardreset == sata_sff_hardreset) && !sata_scr_valid (&ap->link)) - hardreset =NULL ; - -ata_do_eh (ap, ap->ops->prereset, softreset, hardreset, - ap->ops->postreset); +ata_std_error_handler (ap); }EXPORT_SYMBOL_GPL (ata_sff_error_handler);include/linux/libata.h +0
-3@@ -1412,9 +1412,6 @@ extern void ata_eh_thaw_port(struct ata_port *ap);extern void ata_eh_qc_complete (struct ata_queued_cmd *qc);extern void ata_eh_qc_retry (struct ata_queued_cmd *qc); -extern void ata_do_eh (struct ata_port *ap,ata_prereset_fn_t prereset, -ata_reset_fn_t softreset,ata_reset_fn_t hardreset, -ata_postreset_fn_t postreset);extern void ata_std_error_handler (struct ata_port *ap);extern void ata_std_sched_eh (struct ata_port *ap);extern void ata_std_end_eh (struct ata_port *ap);
include/linux/libata.h static inline u8 ata_wait_idle(struct ata_port *ap) · line 2153
Quoted 5 lines
include/linux/libata.h +2
-0@@ -2152,6 +2149,14 @@ static inline u8 ata_wait_idle(struct ata_port *ap) return status; } +#else /* CONFIG_ATA_SFF */ +
Looking at this file, we usually don't have newlines after #else, so I would drop this newline.
include/linux/libata.h static inline u8 ata_wait_idle(struct ata_port *ap) · line 2159
Quoted 6 lines
include/linux/libata.h +6
-0+static inline int sata_sff_hardreset (struct ata_link *link,unsigned int *class, +unsigned long deadline) +{ +return -EOPNOTSUPP ; +} +
Looking at this file, we usually don't have newlines before #endif, so I would drop this newline.
Quoted 6 lines
include/linux/libata.h +0
-0#endif /* CONFIG_ATA_SFF */ #endif /* __LINUX_LIBATA_H__ */ Signature · 1 line
2.50.0
Niklas Cassel
maintainerRe: [PATCH v2 2/3] ata: libata-eh: Remove ata_do_eh()
Quoted 1 line · Niklas Cassel
s/operations/operations/
I of course meant:
s/operations/operation/
Quoted 15 lines
And then in:
intata_eh_reset(struct ata_link *link, int classify, bool use_pmp_ops){ ... if (use_pmp_ops) { prereset = ops->prereset; softreset = ops->softreset; hardreset = ops->hardreset; postreset = ops->postreset; } else { prereset = ops->pmp_prereset; softreset = ops->pmp_softreset; hardreset = ops->pmp_hardreset; postreset = ops->pmp_postreset; }
Oops.. these are swapped, but you get the point.
Kind regards,
Niklas
Niklas Cassel
maintainerRe: [PATCH v2 2/3] ata: libata-eh: Remove ata_do_eh()
In reply to Niklas CasselQuoted 11 lines · Niklas Cassel
On Tue, Jul 08, 2025 at 04:36:47PM +0900, Damien Le Moal wrote:
I think it would be nicer to just do:
if ((hardreset ==
sata_std_hardreset||hardreset == sata_sff_hardreset) && !sata_scr_valid(&ap->link)) link->flags |= ATA_LFLAG_NO_HRST;since
ata_eh_reset()will already do
hardreset = NULL; if that flag is set.
Quoted 8 lines
This way, we can also simplify the function signature of
ata_eh_recover()to:int ata_eh_recover(struct
ata_port*ap, booluse_pmp_ops,
structata_link**r_failed_link)and
ata_eh_reset()to:int
ata_eh_reset(struct ata_link *link, int classify, bool use_pmp_ops)
Even nicer would be create a new:
ata_reset_port_ops {
ata_prereset_fn_t prereset;
ata_reset_fn_t softreset;
ata_reset_fn_t hardreset;
ata_postreset_fn_t postreset;
}And then modify:
struct ata_port_operations {
...
struct ata_reset_port_ops *reset_port_ops;
struct ata_reset_port_ops *reset_port_pmp_ops;
...
}and then modify ata_eh_recover() to:
int ata_eh_recover(struct ata_port *ap, struct ata_reset_port_ops *ops,
struct ata_link **r_failed_link)
and then modify ata_eh_reset() to:
int ata_eh_reset(struct ata_link *link, int classify, struct ata_reset_port_ops *ops)
Kind regards,
Niklas