Merge branch 'cxgb4-reduce-memory-footprint-for-collecting-firmware-dump'

Rahul Lakkireddy says:

====================
cxgb4: reduce memory footprint for collecting firmware dump

Firmware dump can be large (upto 2 GB).  In low memory conditions,
ethtool fails to allocate such large memory.  So, use zlib deflate
to compress collected firmware dump.

Patch 1 updates collection logic to use compression.

Patch 2 adds zlib deflate to compress collected firmware dump.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/chelsio/cxgb4/Makefile

@@ -12,3 +12,4 @@ cxgb4-objs := cxgb4_main.o l2t.o smt.o t4_hw.o sge.o clip_tbl.o cxgb4_ethtool.o
 cxgb4-$(CONFIG_CHELSIO_T4_DCB) +=  cxgb4_dcb.o
 cxgb4-$(CONFIG_CHELSIO_T4_FCOE) +=  cxgb4_fcoe.o
 cxgb4-$(CONFIG_DEBUG_FS) += cxgb4_debugfs.o
+cxgb4-$(CONFIG_ZLIB_DEFLATE) += cudbg_zlib.o

drivers/net/ethernet/chelsio/cxgb4/cudbg_common.c

@@ -19,7 +19,8 @@
 #include "cudbg_if.h"
 #include "cudbg_lib_common.h"
 
-int cudbg_get_buff(struct cudbg_buffer *pdbg_buff, u32 size,
+int cudbg_get_buff(struct cudbg_init *pdbg_init,
+		   struct cudbg_buffer *pdbg_buff, u32 size,
 		   struct cudbg_buffer *pin_buff)
 {
 	u32 offset;
@@ -28,17 +29,30 @@ int cudbg_get_buff(struct cudbg_buffer *pdbg_buff, u32 size,
 	if (offset + size > pdbg_buff->size)
 		return CUDBG_STATUS_NO_MEM;
 
+	if (pdbg_init->compress_type != CUDBG_COMPRESSION_NONE) {
+		if (size > pdbg_init->compress_buff_size)
+			return CUDBG_STATUS_NO_MEM;
+
+		pin_buff->data = (char *)pdbg_init->compress_buff;
+		pin_buff->offset = 0;
+		pin_buff->size = size;
+		return 0;
+	}
+
 	pin_buff->data = (char *)pdbg_buff->data + offset;
 	pin_buff->offset = offset;
 	pin_buff->size = size;
-	pdbg_buff->size -= size;
 	return 0;
 }
 
-void cudbg_put_buff(struct cudbg_buffer *pin_buff,
-		    struct cudbg_buffer *pdbg_buff)
+void cudbg_put_buff(struct cudbg_init *pdbg_init,
+		    struct cudbg_buffer *pin_buff)
 {
-	pdbg_buff->size += pin_buff->size;
+	/* Clear compression buffer for re-use */
+	if (pdbg_init->compress_type != CUDBG_COMPRESSION_NONE)
+		memset(pdbg_init->compress_buff, 0,
+		       pdbg_init->compress_buff_size);
+
 	pin_buff->data = NULL;
 	pin_buff->offset = 0;
 	pin_buff->size = 0;

drivers/net/ethernet/chelsio/cxgb4/cudbg_if.h

@@ -87,6 +87,10 @@ struct cudbg_init {
 	struct adapter *adap; /* Pointer to adapter structure */
 	void *outbuf; /* Output buffer */
 	u32 outbuf_size;  /* Output buffer size */
+	u8 compress_type; /* Type of compression to use */
+	void *compress_buff; /* Compression buffer */
+	u32 compress_buff_size; /* Compression buffer size */
+	void *workspace; /* Workspace for zlib */
 };
 
 static inline unsigned int cudbg_mbytes_to_bytes(unsigned int size)

drivers/net/ethernet/chelsio/cxgb4/cudbg_lib_common.h

@@ -26,6 +26,7 @@ enum cudbg_dump_type {
 
 enum cudbg_compression_type {
 	CUDBG_COMPRESSION_NONE = 1,
+	CUDBG_COMPRESSION_ZLIB,
 };
 
 struct cudbg_hdr {
@@ -78,10 +79,11 @@ struct cudbg_error {
 #define CDUMP_MAX_COMP_BUF_SIZE ((64 * 1024) - 1)
 #define CUDBG_CHUNK_SIZE ((CDUMP_MAX_COMP_BUF_SIZE / 1024) * 1024)
 
-int cudbg_get_buff(struct cudbg_buffer *pdbg_buff, u32 size,
+int cudbg_get_buff(struct cudbg_init *pdbg_init,
+		   struct cudbg_buffer *pdbg_buff, u32 size,
+		   struct cudbg_buffer *pin_buff);
+void cudbg_put_buff(struct cudbg_init *pdbg_init,
 		    struct cudbg_buffer *pin_buff);
-void cudbg_put_buff(struct cudbg_buffer *pin_buff,
-		    struct cudbg_buffer *pdbg_buff);
 void cudbg_update_buff(struct cudbg_buffer *pin_buff,
 		       struct cudbg_buffer *pout_buff);
 #endif /* __CUDBG_LIB_COMMON_H__ */

drivers/net/ethernet/chelsio/cxgb4/cudbg_zlib.c

+/*
+ *  Copyright (C) 2018 Chelsio Communications.  All rights reserved.
+ *
+ *  This program is free software; you can redistribute it and/or modify it
+ *  under the terms and conditions of the GNU General Public License,
+ *  version 2, as published by the Free Software Foundation.
+ *
+ *  This program is distributed in the hope it will be useful, but WITHOUT
+ *  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ *  FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ *  more details.
+ *
+ *  The full GNU General Public License is included in this distribution in
+ *  the file called "COPYING".
+ *
+ */
+
+#include <linux/zlib.h>
+
+#include "cxgb4.h"
+#include "cudbg_if.h"
+#include "cudbg_lib_common.h"
+#include "cudbg_zlib.h"
+
+static int cudbg_get_compress_hdr(struct cudbg_buffer *pdbg_buff,
+				  struct cudbg_buffer *pin_buff)
+{
+	if (pdbg_buff->offset + sizeof(struct cudbg_compress_hdr) >
+	    pdbg_buff->size)
+		return CUDBG_STATUS_NO_MEM;
+
+	pin_buff->data = (char *)pdbg_buff->data + pdbg_buff->offset;
+	pin_buff->offset = 0;
+	pin_buff->size = sizeof(struct cudbg_compress_hdr);
+	pdbg_buff->offset += sizeof(struct cudbg_compress_hdr);
+	return 0;
+}
+
+int cudbg_compress_buff(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *pin_buff,
+			struct cudbg_buffer *pout_buff)
+{
+	struct z_stream_s compress_stream = { 0 };
+	struct cudbg_buffer temp_buff = { 0 };
+	struct cudbg_compress_hdr *c_hdr;
+	int rc;
+
+	/* Write compression header to output buffer before compression */
+	rc = cudbg_get_compress_hdr(pout_buff, &temp_buff);
+	if (rc)
+		return rc;
+
+	c_hdr = (struct cudbg_compress_hdr *)temp_buff.data;
+	c_hdr->compress_id = CUDBG_ZLIB_COMPRESS_ID;
+
+	compress_stream.workspace = pdbg_init->workspace;
+	rc = zlib_deflateInit2(&compress_stream, Z_DEFAULT_COMPRESSION,
+			       Z_DEFLATED, CUDBG_ZLIB_WIN_BITS,
+			       CUDBG_ZLIB_MEM_LVL, Z_DEFAULT_STRATEGY);
+	if (rc != Z_OK)
+		return CUDBG_SYSTEM_ERROR;
+
+	compress_stream.next_in = pin_buff->data;
+	compress_stream.avail_in = pin_buff->size;
+	compress_stream.next_out = pout_buff->data + pout_buff->offset;
+	compress_stream.avail_out = pout_buff->size - pout_buff->offset;
+
+	rc = zlib_deflate(&compress_stream, Z_FINISH);
+	if (rc != Z_STREAM_END)
+		return CUDBG_SYSTEM_ERROR;
+
+	rc = zlib_deflateEnd(&compress_stream);
+	if (rc != Z_OK)
+		return CUDBG_SYSTEM_ERROR;
+
+	c_hdr->compress_size = compress_stream.total_out;
+	c_hdr->decompress_size = pin_buff->size;
+	pout_buff->offset += compress_stream.total_out;
+
+	return 0;
+}

drivers/net/ethernet/chelsio/cxgb4/cudbg_zlib.h

+/*
+ *  Copyright (C) 2018 Chelsio Communications.  All rights reserved.
+ *
+ *  This program is free software; you can redistribute it and/or modify it
+ *  under the terms and conditions of the GNU General Public License,
+ *  version 2, as published by the Free Software Foundation.
+ *
+ *  This program is distributed in the hope it will be useful, but WITHOUT
+ *  ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ *  FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ *  more details.
+ *
+ *  The full GNU General Public License is included in this distribution in
+ *  the file called "COPYING".
+ *
+ */
+
+#ifndef __CUDBG_ZLIB_H__
+#define __CUDBG_ZLIB_H__
+
+#include <linux/zlib.h>
+
+#define CUDBG_ZLIB_COMPRESS_ID 17
+#define CUDBG_ZLIB_WIN_BITS 12
+#define CUDBG_ZLIB_MEM_LVL 4
+
+struct cudbg_compress_hdr {
+	u32 compress_id;
+	u64 decompress_size;
+	u64 compress_size;
+	u64 rsvd[32];
+};
+
+static inline int cudbg_get_workspace_size(void)
+{
+#ifdef CONFIG_ZLIB_DEFLATE
+	return zlib_deflate_workspacesize(CUDBG_ZLIB_WIN_BITS,
+					  CUDBG_ZLIB_MEM_LVL);
+#else
+	return 0;
+#endif /* CONFIG_ZLIB_DEFLATE */
+}
+
+#ifndef CONFIG_ZLIB_DEFLATE
+static inline int cudbg_compress_buff(struct cudbg_init *pdbg_init,
+				      struct cudbg_buffer *pin_buff,
+				      struct cudbg_buffer *pout_buff)
+{
+	return 0;
+}
+#else
+int cudbg_compress_buff(struct cudbg_init *pdbg_init,
+			struct cudbg_buffer *pin_buff,
+			struct cudbg_buffer *pout_buff);
+#endif /* CONFIG_ZLIB_DEFLATE */
+#endif /* __CUDBG_ZLIB_H__ */

drivers/net/ethernet/chelsio/cxgb4/cxgb4_cudbg.c

@@ -18,6 +18,7 @@
 #include "t4_regs.h"
 #include "cxgb4.h"
 #include "cxgb4_cudbg.h"
+#include "cudbg_zlib.h"
 
 static const struct cxgb4_collect_entity cxgb4_collect_mem_dump[] = {
 	{ CUDBG_EDC0, cudbg_collect_edc0_meminfo },
@@ -318,6 +319,7 @@ u32 cxgb4_get_dump_length(struct adapter *adap, u32 flag)
 {
 	u32 i, entity;
 	u32 len = 0;
+	u32 wsize;
 
 	if (flag & CXGB4_ETH_DUMP_HW) {
 		for (i = 0; i < ARRAY_SIZE(cxgb4_collect_hw_dump); i++) {
@@ -333,6 +335,11 @@ u32 cxgb4_get_dump_length(struct adapter *adap, u32 flag)
 		}
 	}
 
+	/* If compression is enabled, a smaller destination buffer is enough */
+	wsize = cudbg_get_workspace_size();
+	if (wsize && len > CUDBG_DUMP_BUFF_SIZE)
+		len = CUDBG_DUMP_BUFF_SIZE;
+
 	return len;
 }
 
@@ -341,22 +348,14 @@ static void cxgb4_cudbg_collect_entity(struct cudbg_init *pdbg_init,
 				       const struct cxgb4_collect_entity *e_arr,
 				       u32 arr_size, void *buf, u32 *tot_size)
 {
-	struct adapter *adap = pdbg_init->adap;
 	struct cudbg_error cudbg_err = { 0 };
 	struct cudbg_entity_hdr *entity_hdr;
-	u32 entity_size, i;
-	u32 total_size = 0;
+	u32 i, total_size = 0;
 	int ret;
 
 	for (i = 0; i < arr_size; i++) {
 		const struct cxgb4_collect_entity *e = &e_arr[i];
 
-		/* Skip entities that won't fit in output buffer */
-		entity_size = cxgb4_get_entity_length(adap, e->entity);
-		if (entity_size >
-		    pdbg_init->outbuf_size - *tot_size - total_size)
-			continue;
-
 		entity_hdr = cudbg_get_entity_hdr(buf, e->entity);
 		entity_hdr->entity_type = e->entity;
 		entity_hdr->start_offset = dbg_buff->offset;
@@ -382,6 +381,28 @@ static void cxgb4_cudbg_collect_entity(struct cudbg_init *pdbg_init,
 	*tot_size += total_size;
 }
 
+static int cudbg_alloc_compress_buff(struct cudbg_init *pdbg_init)
+{
+	u32 workspace_size;
+
+	workspace_size = cudbg_get_workspace_size();
+	pdbg_init->compress_buff = vzalloc(CUDBG_COMPRESS_BUFF_SIZE +
+					   workspace_size);
+	if (!pdbg_init->compress_buff)
+		return -ENOMEM;
+
+	pdbg_init->compress_buff_size = CUDBG_COMPRESS_BUFF_SIZE;
+	pdbg_init->workspace = (u8 *)pdbg_init->compress_buff +
+			       CUDBG_COMPRESS_BUFF_SIZE - workspace_size;
+	return 0;
+}
+
+static void cudbg_free_compress_buff(struct cudbg_init *pdbg_init)
+{
+	if (pdbg_init->compress_buff)
+		vfree(pdbg_init->compress_buff);
+}
+
 int cxgb4_cudbg_collect(struct adapter *adap, void *buf, u32 *buf_size,
 			u32 flag)
 {
@@ -389,6 +410,7 @@ int cxgb4_cudbg_collect(struct adapter *adap, void *buf, u32 *buf_size,
 	struct cudbg_buffer dbg_buff = { 0 };
 	u32 size, min_size, total_size = 0;
 	struct cudbg_hdr *cudbg_hdr;
+	int rc;
 
 	size = *buf_size;
 
@@ -408,7 +430,6 @@ int cxgb4_cudbg_collect(struct adapter *adap, void *buf, u32 *buf_size,
 	cudbg_hdr->max_entities = CUDBG_MAX_ENTITY;
 	cudbg_hdr->chip_ver = adap->params.chip;
 	cudbg_hdr->dump_type = CUDBG_DUMP_TYPE_MINI;
-	cudbg_hdr->compress_type = CUDBG_COMPRESSION_NONE;
 
 	min_size = sizeof(struct cudbg_hdr) +
 		   sizeof(struct cudbg_entity_hdr) *
@@ -416,6 +437,24 @@ int cxgb4_cudbg_collect(struct adapter *adap, void *buf, u32 *buf_size,
 	if (size < min_size)
 		return -ENOMEM;
 
+	rc = cudbg_get_workspace_size();
+	if (rc) {
+		/* Zlib available.  So, use zlib deflate */
+		cudbg_init.compress_type = CUDBG_COMPRESSION_ZLIB;
+		rc = cudbg_alloc_compress_buff(&cudbg_init);
+		if (rc) {
+			/* Ignore error and continue without compression. */
+			dev_warn(adap->pdev_dev,
+				 "Fail allocating compression buffer ret: %d.  Continuing without compression.\n",
+				 rc);
+			cudbg_init.compress_type = CUDBG_COMPRESSION_NONE;
+			rc = 0;
+		}
+	} else {
+		cudbg_init.compress_type = CUDBG_COMPRESSION_NONE;
+	}
+
+	cudbg_hdr->compress_type = cudbg_init.compress_type;
 	dbg_buff.offset += min_size;
 	total_size = dbg_buff.offset;
 
@@ -433,7 +472,11 @@ int cxgb4_cudbg_collect(struct adapter *adap, void *buf, u32 *buf_size,
 					   buf,
 					   &total_size);
 
+	cudbg_free_compress_buff(&cudbg_init);
 	cudbg_hdr->data_len = total_size;
+	if (cudbg_init.compress_type != CUDBG_COMPRESSION_NONE)
+		*buf_size = size;
+	else
 		*buf_size = total_size;
 	return 0;
 }

drivers/net/ethernet/chelsio/cxgb4/cxgb4_cudbg.h

@@ -23,6 +23,9 @@
 #include "cudbg_entity.h"
 #include "cudbg_lib.h"
 
+#define CUDBG_DUMP_BUFF_SIZE (32 * 1024 * 1024) /* 32 MB */
+#define CUDBG_COMPRESS_BUFF_SIZE (4 * 1024 * 1024) /* 4 MB */
+
 typedef int (*cudbg_collect_callback_t)(struct cudbg_init *pdbg_init,
 					struct cudbg_buffer *dbg_buff,
 					struct cudbg_error *cudbg_err);